1 2 3 4 5<!doctype html> 6<html lang="en"> 7<head> 8 <meta charset="utf-8" /> 9 <meta name="viewport" content="width=device-width, initial-scale=1" /> 10 <title>ImageMagick - Architecture</title> 11 <meta name="application-name" content="ImageMagick" /> 12 <meta name="description" content="Use ImageMagick® to create, edit, compose, and convert digital images. Resize an image, crop it, change its shades and colors, add captions, and more." /> 13 <meta name="application-url" content="https://imagemagick.org" /> 14 <meta name="generator" content="PHP" /> 15 <meta name="keywords" content="architecture, image processing software" /> 16 <meta name="rating" content="GENERAL" /> 17 <meta name="robots" content="INDEX, FOLLOW" /> 18 <meta name="generator" content="ImageMagick Studio LLC" /> 19 <meta name="author" content="ImageMagick Studio LLC" /> 20 <meta name="revisit-after" content="2 DAYS" /> 21 <meta name="resource-type" content="document" /> 22 <meta name="copyright" content="Copyright (c) 1999-2020 ImageMagick Studio LLC" /> 23 <meta name="distribution" content="Global" /> 24 <meta name="magick-serial" content="P131-S030410-R485315270133-P82224-A6668-G1245-1" /> 25 <meta property='og:url' content='../' /> 26 <meta property='og:title' content='ImageMagick' /> 27 <meta property='og:image' content='../images/logo.png' /> 28 <meta property='og:type' content='website' /> 29 <meta property='og:site_name' content='ImageMagick' /> 30 <meta property='og:description' content="Create, Edit, Compose, or Convert Digital Images" /> 31 <meta name="google-site-verification" content="_bMOCDpkx9ZAzBwb2kF3PRHbfUUdFj2uO8Jd1AXArz4" /> 32 <link href="architecture.html" rel="canonical" /> 33 <link href="../images/wand.png" rel="icon" /> 34 <link href="../images/wand.ico" rel="shortcut icon" /> 35 <link href="assets/magick.css" rel="stylesheet" /> 36</head> 37<body> 38 <nav class="navbar navbar-expand-md navbar-dark bg-dark fixed-top"> 39 <div class="container-fluid"> 40 <a class="navbar-brand" href="../"><img class="d-block" id="icon" alt="ImageMagick" width="32" height="32" src="../images/wand.ico"/></a> 41 <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#magick-navbars" aria-controls="magick-navbars" aria-expanded="false" aria-label="Toggle navigation"> 42 <span class="navbar-toggler-icon"></span> 43 </button> 44 45 <div class="collapse navbar-collapse" id="magick-navbars"> 46 <ul class="navbar-nav me-auto mb-2 mb-md-0"> 47 <li class="nav-item"> 48 <a class="nav-link " href="../www/index.html">Home</a> 49 </li> 50 <li class="nav-item"> 51 <a class="nav-link " href="../www/download.html">Download</a> 52 </li> 53 <li class="nav-item"> 54 <a class="nav-link " href="../www/command-line-tools.html">Tools</a> 55 </li> 56 <li class="nav-item"> 57 <a class="nav-link " href="../www/command-line-processing.html">CLI</a> 58 </li> 59 <li class="nav-item"> 60 <a class="nav-link " href="../www/develop.html">Develop</a> 61 </li> 62 <li class="nav-item"> 63 <a class="nav-link" target="_blank" href="https://github.com/ImageMagick/ImageMagick/discussions">Community</a> 64 </li> 65 <li class="nav-item"> 66 <iframe src="https://github.com/sponsors/ImageMagick/button" title="Sponsor ImageMagick" height="35" width="107" style="border: 0;"></iframe> 67 </li> 68 </ul> 69 <form class="d-flex form-inline" action="search.html"> 70 <input class="form-control me-2" type="text" name="q" placeholder="Search" aria-label="Search"> 71 <button class="btn btn-outline-success" type="submit" name="sa">Search</button> 72 </form> 73 </div> 74 </div> 75 </nav> 76 77 <div class="container"> 78 <script async="async" src="https://localhost/pagead/js/adsbygoogle.js"></script> 79 <ins class="adsbygoogle" 80 style="display:block" 81 data-ad-client="ca-pub-3129977114552745" 82 data-ad-slot="6345125851" 83 data-full-width-responsive="true" 84 data-ad-format="horizontal"></ins> 85 <script> 86 (adsbygoogle = window.adsbygoogle || []).push({}); 87 </script> 88 89 </div> 90 91 <main class="container"> 92 <div class="magick-template"> 93<div class="magick-header"> 94<h1 class="text-center">Architecture</h1> 95<p class="text-center"><a href="architecture.html#cache">The Pixel Cache</a> • <a href="architecture.html#stream">Streaming Pixels</a> • <a href="architecture.html#properties">Image Properties and Profiles</a> • <a href="architecture.html#tera-pixel">Large Image Support</a> • <a href="architecture.html#threads">Threads of Execution</a> • <a href="architecture.html#distributed">Heterogeneous Distributed Processing</a> • <a href="architecture.html#coders">Custom Image Coders</a> • <a href="architecture.html#filters">Custom Image Filters</a></p> 96 97<p class="lead magick-description">The citizens of Oz were quite content with their benefactor, the all-powerful Wizard. They accepted his wisdom and benevolence without ever questioning the who, why, and where of his power. Like the citizens of Oz, if you feel comfortable that ImageMagick can help you convert, edit, or compose your images without knowing what goes on behind the curtain, feel free to skip this section. However, if you want to know more about the software and algorithms behind ImageMagick, read on. To fully benefit from this discussion, you should be comfortable with image nomenclature and be familiar with computer programming.</p> 98 99<h2><a class="anchor" id="overview"></a>Architecture Overview</h2> 100 101<p>An image typically consists of a rectangular region of pixels and metadata. To convert, edit, or compose an image in an efficient manner, we need convenient access to any pixel anywhere within the region (and sometimes outside the region). And in the case of an image sequence, we need access to any pixel of any region of any image in the sequence. However, there are hundreds of image formats such JPEG, TIFF, PNG, GIF, etc., that makes it difficult to access pixels on demand. Within these formats we find differences in:</p> 102 103<ul> 104 <li>colorspace (e.g sRGB, linear RGB, linear GRAY, CMYK, YUV, Lab, etc.)</li> 105 <li>bit depth (.e.g 1, 4, 8, 12, 16, etc.)</li> 106 <li>storage format (e.g. unsigned, signed, float, double, etc.)</li> 107 <li>compression (e.g. uncompressed, RLE, Zip, BZip, etc.)</li> 108 <li>orientation (i.e. top-to-bottom, right-to-left, etc.),</li> 109 <li>layout (.e.g. raw, interspersed with opcodes, etc.)</li> 110</ul> 111 112<p>In addition, some image pixels may require attenuation, some formats permit more than one frame, and some formats contain vector graphics that must first be rasterized (converted from vector to pixels).</p> 113 114<p>An efficient implementation of an image processing algorithm may require we get or set:</p> 115 116<ul> 117 <li>one pixel a time (e.g. pixel at location 10,3)</li> 118 <li>a single scanline (e.g. all pixels from row 4)</li> 119 <li>a few scanlines at once (e.g. pixel rows 4-7)</li> 120 <li>a single column or columns of pixels (e.g. all pixels from column 11)</li> 121 <li>an arbitrary region of pixels from the image (e.g. pixels defined at 10,7 to 10,19)</li> 122 <li>a pixel in random order (e.g. pixel at 14,15 and 640,480)</li> 123 <li>pixels from two different images (e.g. pixel at 5,1 from image 1 and pixel at 5,1 from image 2)</li> 124 <li>pixels outside the boundaries of the image (e.g. pixel at -1,-3)</li> 125 <li>a pixel component that is unsigned (65311) or in a floating-point representation (e.g. 0.17836)</li> 126 <li>a high-dynamic range pixel that can include negative values (e.g. -0.0072973525628) as well as values that exceed the quantum depth (e.g. 65931)</li> 127 <li>one or more pixels simultaneously in different threads of execution</li> 128 <li>all the pixels in memory to take advantage of speed-ups offered by executing in concert across heterogeneous platforms consisting of CPUs, GPUs, and other processors</li> 129 <li>traits associated with each channel to specify whether the pixel channel is copied, updated, or blended</li> 130 <li>masks that define which pixels are eligible to be updated</li> 131 <li>extra channels that benefits the user but otherwise remain untouched by ImageMagick image processing algorithms</li> 132</ul> 133 134<p>Given the varied image formats and image processing requirements, we implemented the ImageMagick <a href="architecture.html#cache">pixel cache</a> to provide convenient sequential or parallel access to any pixel on demand anywhere inside the image region (i.e. <a href="architecture.html#authentic-pixels">authentic pixels</a>) and from any image in a sequence. In addition, the pixel cache permits access to pixels outside the boundaries defined by the image (i.e. <a href="architecture.html#virtual-pixels">virtual pixels</a>).</p> 135 136<p>In addition to pixels, images have a plethora of <a href="architecture.html#properties">image properties and profiles</a>. Properties include the well known attributes such as width, height, depth, and colorspace. An image may have optional properties which might include the image author, a comment, a create date, and others. Some images also include profiles for color management, or EXIF, IPTC, 8BIM, or XMP informational profiles. ImageMagick provides command line options and programming methods to get, set, or view image properties or profiles or apply profiles.</p> 137 138<p>ImageMagick consists of nearly a half million lines of C code and optionally depends on several million lines of code in dependent libraries (e.g. JPEG, PNG, TIFF libraries). Given that, one might expect a huge architecture document. However, a great majority of image processing is simply accessing pixels and its metadata and our simple, elegant, and efficient implementation makes this easy for the ImageMagick developer. We discuss the implementation of the pixel cache and getting and setting image properties and profiles in the next few sections. Next, we discuss using ImageMagick within a <a href="architecture.html#threads">thread</a> of execution. In the final sections, we discuss <a href="architecture.html#coders">image coders</a> to read or write a particular image format followed by a few words on creating a <a href="architecture.html#filters">filter</a> to access or update pixels based on your custom requirements.</p> 139 140<h2><a class="anchor" id="cache"></a>The Pixel Cache</h2> 141 142<p>The ImageMagick pixel cache is a repository for image pixels with up to 32 channels. The channels are stored contiguously at the depth specified when ImageMagick was built. The channel depths are 8 bits-per-pixel component for the Q8 version of ImageMagick, 16 bits-per-pixel component for the Q16 version, and 32 bits-per-pixel component for the Q32 version. By default pixel components are 32-bit floating-bit <a href="../www/high-dynamic-range.html">high dynamic-range</a> quantities. The channels can hold any value but typically contain red, green, blue, and alpha intensities or cyan, magenta, yellow, black and alpha intensities. A channel might contain the colormap indexes for colormapped images or the black channel for CMYK images. The pixel cache storage may be heap memory, disk-backed memory mapped, or on disk. The pixel cache is reference-counted. Only the cache properties are copied when the cache is cloned. The cache pixels are subsequently copied only when you signal your intention to update any of the pixels.</p> 143 144<h3>Create the Pixel Cache</h3> 145 146<p>The pixel cache is associated with an image when it is created and it is initialized when you try to get or put pixels. Here are three common methods to associate a pixel cache with an image:</p> 147 148<dl> 149<dt class="col-md-8">Create an image canvas initialized to the background color:</dt><br/> 150<dd class="col-md-8"><pre class="highlight"><code>image=AllocateImage(image_info); 151if (SetImageExtent(image,640,480) == MagickFalse) 152 { /* an exception was thrown */ } 153(void) QueryMagickColor("red",&image->background_color,&image->exception); 154SetImageBackgroundColor(image); 155</code></pre></dd> 156 157<dt class="col-md-8">Create an image from a JPEG image on disk:</dt><br/> 158<dd class="col-md-8"><pre class="highlight"><code>(void) strcpy(image_info->filename,"image.jpg"): 159image=ReadImage(image_info,exception); 160if (image == (Image *) NULL) 161 { /* an exception was thrown */ } 162</code></pre></dd> 163<dt class="col-md-8">Create an image from a memory based image:</dt><br/> 164<dd class="col-md-8"><pre class="highlight"><code>image=BlobToImage(blob_info,blob,extent,exception); 165if (image == (Image *) NULL) 166 { /* an exception was thrown */ } 167</code></pre></dd> 168</dl> 169 170<p>In our discussion of the pixel cache, we use the <a href="magick-core.html">MagickCore API</a> to illustrate our points, however, the principles are the same for other program interfaces to ImageMagick.</p> 171 172<p>When the pixel cache is initialized, pixels are scaled from whatever bit depth they originated from to that required by the pixel cache. For example, a 1-channel 1-bit monochrome PBM image is scaled to 8-bit gray image, if you are using the Q8 version of ImageMagick, and 16-bit RGBA for the Q16 version. You can determine which version you have with the <a class="text-nowrap" href="../www/command-line-options.html#version">-version</a> option: </p> 173 174<pre><span class="crtprompt">$ </span><span class='crtin'>identify -version</span><span class='crtout'><br/></span><span class="crtprompt">$ </span><span class='crtin'>Version: ImageMagick 7.0.10-62 2021-01-30 Q16 https://imagemagick.org</span></pre> 175<p>As you can see, the convenience of the pixel cache sometimes comes with a trade-off in storage (e.g. storing a 1-bit monochrome image as 16-bit is wasteful) and speed (i.e. storing the entire image in memory is generally slower than accessing one scanline of pixels at a time). In most cases, the benefits of the pixel cache typically outweigh any disadvantages.</p> 176 177<h3><a class="anchor" id="authentic-pixels"></a>Access the Pixel Cache</h3> 178 179<p>Once the pixel cache is associated with an image, you typically want to get, update, or put pixels into it. We refer to pixels inside the image region as <a href="architecture.html#authentic-pixels">authentic pixels</a> and outside the region as <a href="architecture.html#virtual-pixels">virtual pixels</a>. Use these methods to access the pixels in the cache:</p> 180<ul> 181 <li><a href="api/cache.html#GetVirtualPixels">GetVirtualPixels()</a>: gets pixels that you do not intend to modify or pixels that lie outside the image region (e.g. pixel @ -1,-3)</li> 182 <li><a href="api/cache.html#GetAuthenticPixels">GetAuthenticPixels()</a>: gets pixels that you intend to modify</li> 183 <li><a href="api/cache.html#QueueAuthenticPixels">QueueAuthenticPixels()</a>: queue pixels that you intend to set</li> 184 <li><a href="api/cache.html#SyncAuthenticPixels">SyncAuthenticPixels()</a>: update the pixel cache with any modified pixels</li> 185</ul> 186 187<p>Here is a typical <a href="magick-core.html">MagickCore</a> code snippet for manipulating pixels in the pixel cache. In our example, we copy pixels from the input image to the output image and decrease the intensity by 10%:</p> 188 189<pre class="pre-scrollable highlight"><code>const Quantum 190 *p; 191 192Quantum 193 *q; 194 195ssize_t 196 x, 197 y; 198 199destination=CloneImage(source,source->columns,source->rows,MagickTrue, 200 exception); 201if (destination == (Image *) NULL) 202 { /* an exception was thrown */ } 203for (y=0; y < (ssize_t) source->rows; y++) 204{ 205 p=GetVirtualPixels(source,0,y,source->columns,1,exception); 206 q=GetAuthenticPixels(destination,0,y,destination->columns,1,exception); 207 if ((p == (const Quantum *) NULL) || (q == (Quantum *) NULL) 208 break; 209 for (x=0; x < (ssize_t) source->columns; x++) 210 { 211 SetPixelRed(image,90*p->red/100,q); 212 SetPixelGreen(image,90*p->green/100,q); 213 SetPixelBlue(image,90*p->blue/100,q); 214 SetPixelAlpha(image,90*p->opacity/100,q); 215 p+=GetPixelChannels(source); 216 q+=GetPixelChannels(destination); 217 } 218 if (SyncAuthenticPixels(destination,exception) == MagickFalse) 219 break; 220} 221if (y < (ssize_t) source->rows) 222 { /* an exception was thrown */ } 223</code></pre> 224 225<p>When we first create the destination image by cloning the source image, the pixel cache pixels are not copied. They are only copied when you signal your intentions to modify or set the pixel cache by calling <a href="api/cache.html#GetAuthenticPixels">GetAuthenticPixels()</a> or <a href="api/cache.html#QueueAuthenticPixels">QueueAuthenticPixels()</a>. Use <a href="api/cache.html#QueueAuthenticPixels">QueueAuthenticPixels()</a> if you want to set new pixel values rather than update existing ones. You could use GetAuthenticPixels() to set pixel values but it is slightly more efficient to use QueueAuthenticPixels() instead. Finally, use <a href="api/cache.html#SyncAuthenticPixels">SyncAuthenticPixels()</a> to ensure any updated pixels are pushed to the pixel cache.</p> 226 227<p>You can associate arbitrary content with each pixel, called <em>meta</em> content. Use <a href="api/cache.html#GetVirtualMetacontent">GetVirtualMetacontent()</a> (to read the content) or <a href="api/cache.html#GetAuthenticMetacontent">GetAuthenticMetacontent()</a> (to update the content) to gain access to this content. For example, to print the metacontent, use:</p> 228 229<pre class="highlight"><code>const void 230 *metacontent; 231 232for (y=0; y < (ssize_t) source->rows; y++) 233{ 234 p=GetVirtualPixels(source,0,y,source->columns,1); 235 if (p == (const Quantum *) NULL) 236 break; 237 metacontent=GetVirtualMetacontent(source); 238 /* print meta content here */ 239} 240if (y < (ssize_t) source->rows) 241 /* an exception was thrown */ 242</code></pre> 243 244<p>The pixel cache manager decides whether to give you direct or indirect access to the image pixels. In some cases the pixels are staged to an intermediate buffer-- and that is why you must call SyncAuthenticPixels() to ensure this buffer is <var>pushed</var> out to the pixel cache to guarantee the corresponding pixels in the cache are updated. For this reason we recommend that you only read or update a scanline or a few scanlines of pixels at a time. However, you can get any rectangular region of pixels you want. GetAuthenticPixels() requires that the region you request is within the bounds of the image area. For a 640 by 480 image, you can get a scanline of 640 pixels at row 479 but if you ask for a scanline at row 480, an exception is returned (rows are numbered starting at 0). GetVirtualPixels() does not have this constraint. For example,</p> 245 246<pre class="highlight"><code>p=GetVirtualPixels(source,-3,-3,source->columns+3,6,exception); 247</code></pre> 248 249<p>gives you the pixels you asked for without complaint, even though some are not within the confines of the image region.</p> 250 251<h3><a class="anchor" id="virtual-pixels"></a>Virtual Pixels</h3> 252 253<p>There are a plethora of image processing algorithms that require a neighborhood of pixels about a pixel of interest. The algorithm typically includes a caveat concerning how to handle pixels around the image boundaries, known as edge pixels. With virtual pixels, you do not need to concern yourself about special edge processing other than choosing which virtual pixel method is most appropriate for your algorithm.</p> 254 <p>Access to the virtual pixels are controlled by the <a href="api/cache.html#SetImageVirtualPixelMethod">SetImageVirtualPixelMethod()</a> method from the MagickCore API or the <a class="text-nowrap" href="../www/command-line-options.html#virtual-pixel">-virtual-pixel</a> option from the command line. The methods include:</p> 255 256<dl class="row"> 257<dt class="col-md-4">background</dt> 258<dd class="col-md-8">the area surrounding the image is the background color</dd> 259<dt class="col-md-4">black</dt> 260<dd class="col-md-8">the area surrounding the image is black</dd> 261<dt class="col-md-4">checker-tile</dt> 262<dd class="col-md-8">alternate squares with image and background color</dd> 263<dt class="col-md-4">dither</dt> 264<dd class="col-md-8">non-random 32x32 dithered pattern</dd> 265<dt class="col-md-4">edge</dt> 266<dd class="col-md-8">extend the edge pixel toward infinity (default)</dd> 267<dt class="col-md-4">gray</dt> 268<dd class="col-md-8">the area surrounding the image is gray</dd> 269<dt class="col-md-4">horizontal-tile</dt> 270<dd class="col-md-8">horizontally tile the image, background color above/below</dd> 271<dt class="col-md-4">horizontal-tile-edge</dt> 272<dd class="col-md-8">horizontally tile the image and replicate the side edge pixels</dd> 273<dt class="col-md-4">mirror</dt> 274<dd class="col-md-8">mirror tile the image</dd> 275<dt class="col-md-4">random</dt> 276<dd class="col-md-8">choose a random pixel from the image</dd> 277<dt class="col-md-4">tile</dt> 278<dd class="col-md-8">tile the image</dd> 279<dt class="col-md-4">transparent</dt> 280<dd class="col-md-8">the area surrounding the image is transparent blackness</dd> 281<dt class="col-md-4">vertical-tile</dt> 282<dd class="col-md-8">vertically tile the image, sides are background color</dd> 283<dt class="col-md-4">vertical-tile-edge</dt> 284<dd class="col-md-8">vertically tile the image and replicate the side edge pixels</dd> 285<dt class="col-md-4">white</dt> 286<dd class="col-md-8">the area surrounding the image is white</dd> 287</dl> 288 289 290<h3>Cache Storage and Resource Requirements</h3> 291 292<p>Recall that this simple and elegant design of the ImageMagick pixel cache comes at a cost in terms of storage and processing speed. The pixel cache storage requirements scales with the area of the image and the bit depth of the pixel components. For example, if we have a 640 by 480 image and we are using the non-HDRI Q16 version of ImageMagick, the pixel cache consumes image <var>width * height * bit-depth / 8 * channels</var> bytes or approximately 2.3 mebibytes (i.e. 640 * 480 * 2 * 4). Not too bad, but what if your image is 25000 by 25000 pixels? The pixel cache requires approximately 4.7 gibibytes of storage. Ouch. ImageMagick accounts for possible huge storage requirements by caching large images to disk rather than memory. Typically the pixel cache is stored in memory using heap memory. If heap memory is exhausted, we create the pixel cache on disk and attempt to memory-map it. If memory-map memory is exhausted, we simply use standard disk I/O. Disk storage is cheap but it is also very slow, upwards of 1000 times slower than memory. We can get some speed improvements, up to 5 times, if we use memory mapping to the disk-based cache. These decisions about storage are made <var>automagically</var> by the pixel cache manager negotiating with the operating system. However, you can influence how the pixel cache manager allocates the pixel cache with <var>cache resource limits</var>. The limits include:</p> 293 294<dl class="row"> 295 <dt class="col-md-4">width</dt> 296 <dd class="col-md-8">maximum width of an image. Exceed this limit and an exception is thrown and processing stops.</dd> 297 <dt class="col-md-4">height</dt> 298 <dd class="col-md-8">maximum height of an image. Exceed this limit and an exception is thrown and processing stops.</dd> 299 <dt class="col-md-4">area</dt> 300 <dd class="col-md-8">maximum area in bytes of any one image that can reside in the pixel cache memory. If this limit is exceeded, the image is automagically cached to disk and optionally memory-mapped.</dd> 301 <dt class="col-md-4">memory</dt> 302 <dd class="col-md-8">maximum amount of memory in bytes to allocate for the pixel cache from the heap.</dd> 303 <dt class="col-md-4">map</dt> 304 <dd class="col-md-8">maximum amount of memory map in bytes to allocate for the pixel cache.</dd> 305 <dt class="col-md-4">disk</dt> 306 <dd class="col-md-8">maximum amount of disk space in bytes permitted for use by the pixel cache. If this limit is exceeded, the pixel cache is not created and a fatal exception is thrown.</dd> 307 <dt class="col-md-4">files</dt> 308 <dd class="col-md-8">maximum number of open pixel cache files. When this limit is exceeded, any subsequent pixels cached to disk are closed and reopened on demand. This behavior permits a large number of images to be accessed simultaneously on disk, but without a speed penalty due to repeated open/close calls.</dd> 309 <dt class="col-md-4">thread</dt> 310 <dd class="col-md-8">maximum number of threads that are permitted to run in parallel.</dd> 311 <dt class="col-md-4">time</dt> 312 <dd class="col-md-8">maximum number of seconds that the process is permitted to execute. Exceed this limit and an exception is thrown and processing stops.</dd> 313</dl> 314 315<p>Note, these limits pertain to the ImageMagick pixel cache. Certain algorithms within ImageMagick do not respect these limits nor does any of the external delegate libraries (e.g. JPEG, TIFF, etc.).</p> 316 317<p>To determine the current setting of these limits, use this command:</p> 318<pre class="highlight">-> identify -list resource 319Resource limits: 320 Width: 100MP 321 Height: 100MP 322 Area: 25.181GB 323 Memory: 11.726GiB 324 Map: 23.452GiB 325 Disk: unlimited 326 File: 768 327 Thread: 12 328 Throttle: 0 329 Time: unlimited 330</pre> 331 332<p>You can set these limits either as a <a href="../www/security-policy.html">security policy</a> (see <a href="https://imagemagick.org/source/policy.xml">policy.xml</a>), with an <a href="resources.html#environment">environment variable</a>, with the <a href="../www/command-line-options.html#limit">-limit</a> command line option, or with the <a href="api/resource.html#SetMagickResourceLimit">SetMagickResourceLimit()</a> MagickCore API method. As an example, our online web interface to ImageMagick, <a href="../MagickStudio/scripts/MagickStudio.cgi">ImageMagick Studio</a>, includes these policy limits to help prevent a denial-of-service:</p> 333<pre class="highlight"><code><policymap> 334 <!-- temporary path must be a preexisting writable directory --> 335 <policy domain="resource" name="temporary-path" value="/tmp"/> 336 <policy domain="resource" name="memory" value="256MiB"/> 337 <policy domain="resource" name="map" value="512MiB"/> 338 <policy domain="resource" name="width" value="8KP"/> 339 <policy domain="resource" name="height" value="8KP"/> 340 <policy domain="resource" name="area" value="16KP"/> 341 <policy domain="resource" name="disk" value="1GiB"/> 342 <policy domain="resource" name="file" value="768"/> 343 <policy domain="resource" name="thread" value="2"/> 344 <policy domain="resource" name="throttle" value="0"/> 345 <policy domain="resource" name="time" value="120"/> 346 <policy domain="resource" name="list-length" value="128"/> 347 <policy domain="system" name="precision" value="6"/> 348 <policy domain="cache" name="shared-secret" stealth="true" value="replace with your secret phrase"/> 349 <policy domain="coder" rights="none" pattern="MVG" /> 350 <policy domain="coder" rights="none" pattern="EPS" /> 351 <policy domain="coder" rights="none" pattern="PS" /> 352 <policy domain="coder" rights="none" pattern="PS2" /> 353 <policy domain="coder" rights="none" pattern="PS3" /> 354 <policy domain="coder" rights="none" pattern="PDF" /> 355 <policy domain="coder" rights="none" pattern="XPS" /> 356 <policy domain="filter" rights="none" pattern="*" /> 357 <policy domain="delegate" rights="none" pattern="HTTPS" /> <!-- prevent 'curl' program from reading HTTPS URL's --> 358 <policy domain="delegate" rights="none" pattern="SHOW" /> 359 <policy domain="delegate" rights="none" pattern="WIN" /> 360 <policy domain="path" rights="none" pattern="@*"/> <!-- indirect reads not permitted --> 361</code></pre> 362<p>Since we process multiple simultaneous sessions, we don't want any one session consuming all the available memory.With this policy, large images are cached to disk. If the image is too large and exceeds the pixel cache disk limit, the program exits. In addition, we place a time limit to prevent any run-away processing tasks. If any one image has a width or height that exceeds 8192 pixels, an exception is thrown and processing stops. As of ImageMagick 7.0.1-8 you can prevent the use of any delegate or all delegates (set the pattern to "*"). Note, prior to this release, use a domain of "coder" to prevent delegate usage (e.g. domain="coder" rights="none" pattern="HTTPS"). The policy also prevents indirect reads. If you want to, for example, read text from a file (e.g. caption:@myCaption.txt), you'll need to remove this policy.</p> 363 364<p>Note, the cache limits are global to each invocation of ImageMagick, meaning if you create several images, the combined resource requirements are compared to the limit to determine the pixel cache storage disposition.</p> 365 366<p>To determine which type and how much resources are consumed by the pixel cache, add the <a href="../www/command-line-options.html#debug">-debug cache</a> option to the command-line:</p> 367<pre class="highlight">-> magick -debug cache logo: -sharpen 3x2 null: 3682016-12-17T13:33:42-05:00 0:00.000 0.000u 7.0.0 Cache magick: cache.c/DestroyPixelCache/1275/Cache 369 destroy 3702016-12-17T13:33:42-05:00 0:00.000 0.000u 7.0.0 Cache magick: cache.c/OpenPixelCache/3834/Cache 371 open LOGO[0] (Heap Memory, 640x480x4 4.688MiB) 3722016-12-17T13:33:42-05:00 0:00.010 0.000u 7.0.0 Cache magick: cache.c/OpenPixelCache/3834/Cache 373 open LOGO[0] (Heap Memory, 640x480x3 3.516MiB) 3742016-12-17T13:33:42-05:00 0:00.010 0.000u 7.0.0 Cache magick: cache.c/ClonePixelCachePixels/1044/Cache 375 Memory => Memory 3762016-12-17T13:33:42-05:00 0:00.020 0.010u 7.0.0 Cache magick: cache.c/ClonePixelCachePixels/1044/Cache 377 Memory => Memory 3782016-12-17T13:33:42-05:00 0:00.020 0.010u 7.0.0 Cache magick: cache.c/OpenPixelCache/3834/Cache 379 open LOGO[0] (Heap Memory, 640x480x3 3.516MiB) 3802016-12-17T13:33:42-05:00 0:00.050 0.100u 7.0.0 Cache magick: cache.c/DestroyPixelCache/1275/Cache 381 destroy LOGO[0] 3822016-12-17T13:33:42-05:00 0:00.050 0.100u 7.0.0 Cache magick: cache.c/DestroyPixelCache/1275/Cache 383 destroy LOGO[0] 384</pre> 385<p>This command utilizes a pixel cache in memory. The logo consumed 4.688MiB and after it was sharpened, 3.516MiB.</p> 386 387 388<h3>Distributed Pixel Cache</h3> 389<p>A distributed pixel cache is an extension of the traditional pixel cache available on a single host. The distributed pixel cache may span multiple servers so that it can grow in size and transactional capacity to support very large images. Start up the pixel cache server on one or more machines. When you read or operate on an image and the local pixel cache resources are exhausted, ImageMagick contacts one or more of these remote pixel servers to store or retrieve pixels. The distributed pixel cache relies on network bandwidth to marshal pixels to and from the remote server. As such, it will likely be significantly slower than a pixel cache utilizing local storage (e.g. memory, disk, etc.).</p> 390<pre class="highlight"><code>magick -distribute-cache 6668 & // start on 192.168.100.50 391magick -define registry:cache:hosts=192.168.100.50:6668 myimage.jpg -sharpen 5x2 mimage.png 392</code></pre> 393 394<h3>Cache Views</h3> 395 396<p>GetVirtualPixels(), GetAuthenticPixels(), QueueAuthenticPixels(), and SyncAuthenticPixels(), from the MagickCore API, can only deal with one pixel cache area per image at a time. Suppose you want to access the first and last scanline from the same image at the same time? The solution is to use a <var>cache view</var>. A cache view permits you to access as many areas simultaneously in the pixel cache as you require. The cache view <a href="api/cache-view.html">methods</a> are analogous to the previous methods except you must first open a view and close it when you are finished with it. Here is a snippet of MagickCore code that permits us to access the first and last pixel row of the image simultaneously:</p> 397<pre class="pre-scrollable highlight"><code>CacheView 398 *view_1, 399 *view_2; 400 401view_1=AcquireVirtualCacheView(source,exception); 402view_2=AcquireVirtualCacheView(source,exception); 403for (y=0; y < (ssize_t) source->rows; y++) 404{ 405 u=GetCacheViewVirtualPixels(view_1,0,y,source->columns,1,exception); 406 v=GetCacheViewVirtualPixels(view_2,0,source->rows-y-1,source->columns,1,exception); 407 if ((u == (const Quantum *) NULL) || (v == (const Quantum *) NULL)) 408 break; 409 for (x=0; x < (ssize_t) source->columns; x++) 410 { 411 /* do something with u & v here */ 412 } 413} 414view_2=DestroyCacheView(view_2); 415view_1=DestroyCacheView(view_1); 416if (y < (ssize_t) source->rows) 417 { /* an exception was thrown */ } 418</code></pre> 419 420<h3>Magick Persistent Cache Format</h3> 421 422<p>Recall that each image format is decoded by ImageMagick and the pixels are deposited in the pixel cache. If you write an image, the pixels are read from the pixel cache and encoded as required by the format you are writing (e.g. GIF, PNG, etc.). The Magick Persistent Cache (MPC) format is designed to eliminate the overhead of decoding and encoding pixels to and from an image format. MPC writes two files. One, with the extension <code>.mpc</code>, retains all the properties associated with the image or image sequence (e.g. width, height, colorspace, etc.) and the second, with the extension <code>.cache</code>, is the pixel cache in the native raw format. When reading an MPC image file, ImageMagick reads the image properties and memory maps the pixel cache on disk eliminating the need for decoding the image pixels. The tradeoff is in disk space. MPC is generally larger in file size than most other image formats.</p> 423<p>The most efficient use of MPC image files is a write-once, read-many-times pattern. For example, your workflow requires extracting random blocks of pixels from the source image. Rather than re-reading and possibly decompressing the source image each time, we use MPC and map the image directly to memory.</p> 424 425<h3>Best Practices</h3> 426 427<p>Although you can request any pixel from the pixel cache, any block of pixels, any scanline, multiple scanlines, any row, or multiple rows with the GetVirtualPixels(), GetAuthenticPixels(), QueueAuthenticPixels, GetCacheViewVirtualPixels(), GetCacheViewAuthenticPixels(), and QueueCacheViewAuthenticPixels() methods, ImageMagick is optimized to return a few pixels or a few pixels rows at time. There are additional optimizations if you request a single scanline or a few scanlines at a time. These methods also permit random access to the pixel cache, however, ImageMagick is optimized for sequential access. Although you can access scanlines of pixels sequentially from the last row of the image to the first, you may get a performance boost if you access scanlines from the first row of the image to the last, in sequential order.</p> 428 429<p>You can get, modify, or set pixels in row or column order. However, it is more efficient to access the pixels by row rather than by column.</p> 430 431<p>If you update pixels returned from GetAuthenticPixels() or GetCacheViewAuthenticPixels(), don't forget to call SyncAuthenticPixels() or SyncCacheViewAuthenticPixels() respectively to ensure your changes are synchronized with the pixel cache.</p> 432 433<p>Use QueueAuthenticPixels() or QueueCacheViewAuthenticPixels() if you are setting an initial pixel value. The GetAuthenticPixels() or GetCacheViewAuthenticPixels() method reads pixels from the cache and if you are setting an initial pixel value, this read is unnecessary. Don't forget to call SyncAuthenticPixels() or SyncCacheViewAuthenticPixels() respectively to push any pixel changes to the pixel cache.</p> 434 435<p>GetVirtualPixels(), GetAuthenticPixels(), QueueAuthenticPixels(), and SyncAuthenticPixels() are slightly more efficient than their cache view counter-parts. However, cache views are required if you need access to more than one region of the image simultaneously or if more than one <a href="architecture.html#threads">thread of execution</a> is accessing the image.</p> 436 437<p>You can request pixels outside the bounds of the image with GetVirtualPixels() or GetCacheViewVirtualPixels(), however, it is more efficient to request pixels within the confines of the image region.</p> 438 439<p>Although you can force the pixel cache to disk using appropriate resource limits, disk access can be upwards of 1000 times slower than memory access. For fast, efficient, access to the pixel cache, try to keep the pixel cache in heap memory.</p> 440 441<p>The ImageMagick Q16 version of ImageMagick permits you to read and write 16 bit images without scaling but the pixel cache consumes twice as many resources as the Q8 version. If your system has constrained memory or disk resources, consider the Q8 version of ImageMagick. In addition, the Q8 version typically executes faster than the Q16 version.</p> 442 443<p>A great majority of image formats and algorithms restrict themselves to a fixed range of pixel values from 0 to some maximum value, for example, the Q16 version of ImageMagick permit intensities from 0 to 65535. High dynamic-range imaging (HDRI), however, permits a far greater dynamic range of exposures (i.e. a large difference between light and dark areas) than standard digital imaging techniques. HDRI accurately represents the wide range of intensity levels found in real scenes ranging from the brightest direct sunlight to the deepest darkest shadows. Enable <a href="../www/high-dynamic-range.html">HDRI</a> at ImageMagick build time to deal with high dynamic-range images, but be mindful that each pixel component is a 32-bit floating point value. In addition, pixel values are not clamped by default so some algorithms may have unexpected results due to out-of-band pixel values than the non-HDRI version.</p> 444 445<p>If you are dealing with large images, make sure the pixel cache is written to a disk area with plenty of free space. Under Unix, this is typically <code>/tmp</code> and for Windows, <code>c:/temp</code>. You can tell ImageMagick to write the pixel cache to an alternate location and conserve memory with these options:</p> 446<pre class="highlight"><code>magick -limit memory 2GB -limit map 4GB -define registry:temporary-path=/data/tmp ... 447</code></pre> 448 449<p>Set global resource limits for your environment in the <code>policy.xml</code> configuration file.</p> 450 451<p>If you plan on processing the same image many times, consider the MPC format. Reading a MPC image has near-zero overhead because its in the native pixel cache format eliminating the need for decoding the image pixels. Here is an example:</p> 452<pre class="highlight"><code>magick image.tif image.mpc 453magick image.mpc -crop 100x100+0+0 +repage 1.png 454magick image.mpc -crop 100x100+100+0 +repage 2.png 455magick image.mpc -crop 100x100+200+0 +repage 3.png 456</code></pre> 457 458<p>MPC is ideal for web sites. It reduces the overhead of reading and writing an image. We use it exclusively at our <a href="../MagickStudio/scripts/MagickStudio.cgi">online image studio</a>.</p> 459 460<h2><a class="anchor" id="stream"></a>Streaming Pixels</h2> 461 462<p>ImageMagick provides for streaming pixels as they are read from or written to an image. This has several advantages over the pixel cache. The time and resources consumed by the pixel cache scale with the area of an image, whereas the pixel stream resources scale with the width of an image. The disadvantage is the pixels must be consumed as they are streamed so there is no persistence.</p> 463 464<p>Use <a href="api/stream.html#ReadStream">ReadStream()</a> or <a href="api/stream.html#WriteStream">WriteStream()</a> with an appropriate callback method in your MagickCore program to consume the pixels as they are streaming. Here's an abbreviated example of using ReadStream:</p> 465<pre class="pre-scrollable highlight"><code>static size_t StreamPixels(const Image *image,const void *pixels,const size_t columns) 466{ 467 register const Quantum 468 *p; 469 470 MyData 471 *my_data; 472 473 my_data=(MyData *) image->client_data; 474 p=(Quantum *) pixels; 475 if (p != (const Quantum *) NULL) 476 { 477 /* process pixels here */ 478 } 479 return(columns); 480} 481 482... 483 484/* invoke the pixel stream here */ 485image_info->client_data=(void *) MyData; 486image=ReadStream(image_info,&StreamPixels,exception); 487</code></pre> 488 489<p>We also provide a lightweight tool, <a href="stream.html">stream</a>, to stream one or more pixel components of the image or portion of the image to your choice of storage formats. It writes the pixel components as they are read from the input image a row at a time making <a href="stream.html">stream</a> desirable when working with large images or when you require raw pixel components. A majority of the image formats stream pixels (red, green, and blue) from left to right and top to bottom. However, a few formats do not support this common ordering (e.g. the PSD format).</p> 490 491<h2><a class="anchor" id="properties"></a>Image Properties and Profiles</h2> 492 493<p>Images have metadata associated with them in the form of properties (e.g. width, height, description, etc.) and profiles (e.g. EXIF, IPTC, color management). ImageMagick provides convenient methods to get, set, or update image properties and get, set, update, or apply profiles. Some of the more popular image properties are associated with the Image structure in the MagickCore API. For example:</p> 494<pre class="highlight"><code>(void) printf("image width: %lu, height: %lu\n",image->columns,image->rows); 495</code></pre> 496 497<p>For a great majority of image properties, such as an image comment or description, we use the <a href="api/property.html#GetImageProperty">GetImageProperty()</a> and <a href="api/property.html#SetImageProperty">SetImageProperty()</a> methods. Here we set a property and fetch it right back:</p> 498<pre class="highlight"><code>const char 499 *comment; 500 501(void) SetImageProperty(image,"comment","This space for rent"); 502comment=GetImageProperty(image,"comment"); 503if (comment == (const char *) NULL) 504 (void) printf("Image comment: %s\n",comment); 505</code></pre> 506 507<p>ImageMagick supports artifacts with the GetImageArtifact() and SetImageArtifact() methods. Artifacts are stealth properties that are not exported to image formats (e.g. PNG).</p> 508 509<p>Image profiles are handled with <a href="api/profile.html#GetImageProfile">GetImageProfile()</a>, <a href="api/profile.html#SetImageProfile">SetImageProfile()</a>, and <a href="api/profile.html#ProfileImage">ProfileImage()</a> methods. Here we set a profile and fetch it right back:</p> 510<pre class="highlight"><code>StringInfo 511 *profile; 512 513profile=AcquireStringInfo(length); 514SetStringInfoDatum(profile,my_exif_profile); 515(void) SetImageProfile(image,"EXIF",profile); 516DestroyStringInfo(profile); 517profile=GetImageProfile(image,"EXIF"); 518if (profile != (StringInfo *) NULL) 519 (void) PrintStringInfo(stdout,"EXIF",profile); 520</code></pre> 521 522<h2><a class="anchor" id="tera-pixel"></a>Large Image Support</h2> 523<p>ImageMagick can read, process, or write mega-, giga-, or tera-pixel image sizes. An image width or height can range from 1 to 2 giga-pixels on a 32 bit OS (up to 2147483647 rows/columns) and up to 9 exa-pixels on a 64-bit OS (up to 9223372036854775807 rows/columns). Note, that some image formats have restrictions on image size. For example, Photoshop images are limited to 300,000 pixels for width or height. Here we resize an image to a quarter million pixels square:</p> 524<pre class="highlight"><code>magick logo: -resize 250000x250000 logo.miff 525</code></pre> 526 527<p>For large images, memory resources will likely be exhausted and ImageMagick will instead create a pixel cache on disk. Make sure you have plenty of temporary disk space. If your default temporary disk partition is too small, tell ImageMagick to use another partition with plenty of free space. For example:</p> 528<pre class="highlight"><code>magick -define registry:temporary-path=/data/tmp logo: \ <br/> -resize 250000x250000 logo.miff 529</code></pre> 530 531<p>To ensure large images do not consume all the memory on your system, force the image pixels to memory-mapped disk with resource limits:</p> 532<pre class="highlight"><code>magick -define registry:temporary-path=/data/tmp -limit memory 16mb \ 533 logo: -resize 250000x250000 logo.miff 534</code></pre> 535 536<p>Here we force all image pixels to disk:</p> 537<pre class="highlight"><code>magick -define registry:temporary-path=/data/tmp -limit area 0 \ 538 logo: -resize 250000x250000 logo.miff 539</code></pre> 540 541<p>Caching pixels to disk is about 1000 times slower than memory. Expect long run times when processing large images on disk with ImageMagick. You can monitor progress with this command:</p> 542<pre class="highlight"><code>magick -monitor -limit memory 2GiB -limit map 4GiB -define registry:temporary-path=/data/tmp \ 543 logo: -resize 250000x250000 logo.miff 544</code></pre> 545 546<p>For really large images, or if there is limited resources on your host, you can utilize a distributed pixel cache on one or more remote hosts:</p> 547<pre class="highlight"><code>magick -distribute-cache 6668 & // start on 192.168.100.50 548magick -distribute-cache 6668 & // start on 192.168.100.51 549magick -limit memory 2mb -limit map 2mb -limit disk 2gb \ 550 -define registry:cache:hosts=192.168.100.50:6668,192.168.100.51:6668 \ 551 myhugeimage.jpg -sharpen 5x2 myhugeimage.png 552</code></pre> 553<p>Due to network latency, expect a substantial slow-down in processing your workflow.</p> 554 555<h2><a class="anchor" id="threads"></a>Threads of Execution</h2> 556 557<p>Many of ImageMagick's internal algorithms are threaded to take advantage of speed-ups offered by the multicore processor chips. However, you are welcome to use ImageMagick algorithms in your threads of execution with the exception of the MagickCore's GetVirtualPixels(), GetAuthenticPixels(), QueueAuthenticPixels(), or SyncAuthenticPixels() pixel cache methods. These methods are intended for one thread of execution only with the exception of an OpenMP parallel section. To access the pixel cache with more than one thread of execution, use a cache view. We do this for the <a href="api/composite.html#CompositeImage">CompositeImage()</a> method, for example. Suppose we want to composite a single source image over a different destination image in each thread of execution. If we use GetVirtualPixels(), the results are unpredictable because multiple threads would likely be asking for different areas of the pixel cache simultaneously. Instead we use GetCacheViewVirtualPixels() which creates a unique view for each thread of execution ensuring our program behaves properly regardless of how many threads are invoked. The other program interfaces, such as the <a href="magick-wand.html">MagickWand API</a>, are completely thread safe so there are no special precautions for threads of execution.</p> 558 559<p>Here is an MagickCore code snippet that takes advantage of threads of execution with the <a href="openmp.html">OpenMP</a> programming paradigm:</p> 560<pre class="pre-scrollable highlight"><code>CacheView 561 *image_view; 562 563MagickBooleanType 564 status; 565 566ssize_t 567 y; 568 569status=MagickTrue; 570image_view=AcquireVirtualCacheView(image,exception); 571#pragma omp parallel for schedule(static,4) shared(status) 572for (y=0; y < (ssize_t) image->rows; y++) 573{ 574 register Quantum 575 *q; 576 577 register ssize_t 578 x; 579 580 register void 581 *metacontent; 582 583 if (status == MagickFalse) 584 continue; 585 q=GetCacheViewAuthenticPixels(image_view,0,y,image->columns,1,exception); 586 if (q == (Quantum *) NULL) 587 { 588 status=MagickFalse; 589 continue; 590 } 591 metacontent=GetCacheViewAuthenticMetacontent(image_view); 592 for (x=0; x < (ssize_t) image->columns; x++) 593 { 594 SetPixelRed(image,...,q); 595 SetPixelGreen(image,...,q); 596 SetPixelBlue(image,...,q); 597 SetPixelAlpha(image,...,q); 598 if (metacontent != NULL) 599 metacontent[indexes+x]=...; 600 q+=GetPixelChannels(image); 601 } 602 if (SyncCacheViewAuthenticPixels(image_view,exception) == MagickFalse) 603 status=MagickFalse; 604} 605image_view=DestroyCacheView(image_view); 606if (status == MagickFalse) 607 perror("something went wrong"); 608</code></pre> 609 610<p>This code snippet converts an uncompressed Windows bitmap to a Magick++ image:</p> 611<pre class="pre-scrollable highlight"><code>#include "Magick++.h" 612#include <assert.h> 613#include "omp.h" 614 615void ConvertBMPToImage(const BITMAPINFOHEADER *bmp_info, 616 const unsigned char *restrict pixels,Magick::Image *image) 617{ 618 /* 619 Prepare the image so that we can modify the pixels directly. 620 */ 621 assert(bmp_info->biCompression == BI_RGB); 622 assert(bmp_info->biWidth == image->columns()); 623 assert(abs(bmp_info->biHeight) == image->rows()); 624 image->modifyImage(); 625 if (bmp_info->biBitCount == 24) 626 image->type(MagickCore::TrueColorType); 627 else 628 image->type(MagickCore::TrueColorMatteType); 629 register unsigned int bytes_per_row=bmp_info->biWidth*bmp_info->biBitCount/8; 630 if (bytes_per_row % 4 != 0) { 631 bytes_per_row=bytes_per_row+(4-bytes_per_row % 4); // divisible by 4. 632 } 633 /* 634 Copy all pixel data, row by row. 635 */ 636 #pragma omp parallel for 637 for (int y=0; y < int(image->rows()); y++) 638 { 639 int 640 row; 641 642 register const unsigned char 643 *restrict p; 644 645 register MagickCore::Quantum 646 *restrict q; 647 648 row=(bmp_info->biHeight > 0) ? (image->rows()-y-1) : y; 649 p=pixels+row*bytes_per_row; 650 q=image->setPixels(0,y,image->columns(),1); 651 for (int x=0; x < int(image->columns()); x++) 652 { 653 SetPixelBlue(image,p[0],q); 654 SetPixelGreen(image,p[1],q); 655 SetPixelRed(image,p[2],q); 656 if (bmp_info->biBitCount == 32) { 657 SetPixelAlpha(image,p[3],q); 658 } 659 q+=GetPixelChannels(image); 660 p+=bmp_info->biBitCount/8; 661 } 662 image->syncPixels(); // sync pixels to pixel cache. 663 } 664 return; 665}</code></pre> 666 667<p>If you call the ImageMagick API from your OpenMP-enabled application and you intend to dynamically increase the number of threads available in subsequent parallel regions, be sure to perform the increase <var>before</var> you call the API otherwise ImageMagick may fault.</p> 668 669<p><a href="api/wand-view.html">MagickWand</a> supports wand views. A view iterates over the entire, or portion, of the image in parallel and for each row of pixels, it invokes a callback method you provide. This limits most of your parallel programming activity to just that one module. There are similar methods in <a href="api/image-view.html">MagickCore</a>. For an example, see the same sigmoidal contrast algorithm implemented in both <a href="magick-wand.html#wand-view">MagickWand</a> and <a href="magick-core.html#image-view">MagickCore</a>.</p> 670 671<p>In most circumstances, the default number of threads is set to the number of processor cores on your system for optimal performance. However, if your system is hyperthreaded or if you are running on a virtual host and only a subset of the processors are available to your server instance, you might get an increase in performance by setting the thread <a href="resources.html#configure">policy</a> or the <a href="resources.html#environment">MAGICK_THREAD_LIMIT</a> environment variable. For example, your virtual host has 8 processors but only 2 are assigned to your server instance. The default of 8 threads can cause severe performance problems. One solution is to limit the number of threads to the available processors in your <a href="https://imagemagick.org/source/policy.xml">policy.xml</a> configuration file:</p> 672<pre class="highlight"><code><policy domain="resource" name="thread" value="2"/> 673</code></pre> 674 675<p>Or suppose your 12 core hyperthreaded computer defaults to 24 threads. Set the MAGICK_THREAD_LIMIT environment variable and you will likely get improved performance:</p> 676<pre class="highlight"><code>export MAGICK_THREAD_LIMIT=12 677</code></pre> 678 679<p>The OpenMP committee has not defined the behavior of mixing OpenMP with other threading models such as Posix threads. However, using modern releases of Linux, OpenMP and Posix threads appear to interoperate without complaint. If you want to use Posix threads from a program module that calls one of the ImageMagick application programming interfaces (e.g. MagickCore, MagickWand, Magick++, etc.) from Mac OS X or an older Linux release, you may need to disable OpenMP support within ImageMagick. Add the <code>--disable-openmp</code> option to the configure script command line and rebuild and reinstall ImageMagick.</p> 680 681<p>You can further increase performance by reducing lock contention with the <a href="http://goog-perftools.sourceforge.net/doc/tcmalloc.html">tcmalloc</a> memory allocation library. To enable, add <code>--with-tcmalloc</code> to the <code>configure</code> command-line when you build ImageMagick.</p> 682 683<h5>Threading Performance</h5> 684<p>It can be difficult to predict behavior in a parallel environment. Performance might depend on a number of factors including the compiler, the version of the OpenMP library, the processor type, the number of cores, the amount of memory, whether hyperthreading is enabled, the mix of applications that are executing concurrently with ImageMagick, or the particular image-processing algorithm you utilize. The only way to be certain of optimal performance, in terms of the number of threads, is to benchmark. ImageMagick includes progressive threading when benchmarking a command and returns the elapsed time and efficiency for one or more threads. This can help you identify how many threads is the most efficient in your environment. For this benchmark we sharpen a 1920x1080 image of a model 10 times with 1 to 12 threads:</p> 685<pre class="highlight">-> magick -bench 10 model.png -sharpen 5x2 null: 686Performance[1]: 10i 1.135ips 1.000e 8.760u 0:08.810 687Performance[2]: 10i 2.020ips 0.640e 9.190u 0:04.950 688Performance[3]: 10i 2.786ips 0.710e 9.400u 0:03.590 689Performance[4]: 10i 3.378ips 0.749e 9.580u 0:02.960 690Performance[5]: 10i 4.032ips 0.780e 9.580u 0:02.480 691Performance[6]: 10i 4.566ips 0.801e 9.640u 0:02.190 692Performance[7]: 10i 3.788ips 0.769e 10.980u 0:02.640 693Performance[8]: 10i 4.115ips 0.784e 12.030u 0:02.430 694Performance[9]: 10i 4.484ips 0.798e 12.860u 0:02.230 695Performance[10]: 10i 4.274ips 0.790e 14.830u 0:02.340 696Performance[11]: 10i 4.348ips 0.793e 16.500u 0:02.300 697Performance[12]: 10i 4.525ips 0.799e 18.320u 0:02.210 698</pre> 699<p>The sweet spot for this example is 6 threads. This makes sense since there are 6 physical cores. The other 6 are hyperthreads. It appears that sharpening does not benefit from hyperthreading.</p> 700<p>In certain cases, it might be optimal to set the number of threads to 1 or to disable OpenMP completely with the <a href="resources.html#environment">MAGICK_THREAD_LIMIT</a> environment variable, <a href="../www/command-line-options.html#limit">-limit</a> command line option, or the <a href="resources.html#configure">policy.xml</a> configuration file.</p> 701 702<h2><a class="anchor" id="distributed"></a>Heterogeneous Distributed Processing</h2> 703<p>ImageMagick includes support for heterogeneous distributed processing with the <a href="http://en.wikipedia.org/wiki/OpenCL">OpenCL</a> framework. OpenCL kernels within ImageMagick permit image processing algorithms to execute across heterogeneous platforms consisting of CPUs, GPUs, and other processors. Depending on your platform, speed-ups can be an order of magnitude faster than the traditional single CPU.</p> 704 705<p>First verify that your version of ImageMagick includes support for the OpenCL feature:</p> 706<pre class="highlight"><code>magick identify -version 707Features: DPC Cipher Modules OpenCL OpenMP(4.5) 708</code></pre> 709 710<p>If so, run this command to realize a significant speed-up for image convolution:</p> 711 712<pre class="highlight"><code>magick image.png -convolve '-1, -1, -1, -1, 9, -1, -1, -1, -1' convolve.png 713</code></pre> 714 715<p>If an accelerator is not available or if the accelerator fails to respond, ImageMagick reverts to the non-accelerated convolution algorithm.</p> 716 717<p>Here is an example OpenCL kernel that convolves an image:</p> 718<pre class="pre-scrollable highlight"><code>static inline long ClampToCanvas(const long offset,const ulong range) 719{ 720 if (offset < 0L) 721 return(0L); 722 if (offset >= range) 723 return((long) (range-1L)); 724 return(offset); 725} 726 727static inline CLQuantum ClampToQuantum(const float value) 728{ 729 if (value < 0.0) 730 return((CLQuantum) 0); 731 if (value >= (float) QuantumRange) 732 return((CLQuantum) QuantumRange); 733 return((CLQuantum) (value+0.5)); 734} 735 736__kernel void Convolve(const __global CLPixelType *source,__constant float *filter, 737 const ulong width,const ulong height,__global CLPixelType *destination) 738{ 739 const ulong columns = get_global_size(0); 740 const ulong rows = get_global_size(1); 741 742 const long x = get_global_id(0); 743 const long y = get_global_id(1); 744 745 const float scale = (1.0/QuantumRange); 746 const long mid_width = (width-1)/2; 747 const long mid_height = (height-1)/2; 748 float4 sum = { 0.0, 0.0, 0.0, 0.0 }; 749 float gamma = 0.0; 750 register ulong i = 0; 751 752 for (long v=(-mid_height); v <= mid_height; v++) 753 { 754 for (long u=(-mid_width); u <= mid_width; u++) 755 { 756 register const ulong index=ClampToCanvas(y+v,rows)*columns+ClampToCanvas(x+u, 757 columns); 758 const float alpha=scale*(QuantumRange-source[index].w); 759 sum.x+=alpha*filter[i]*source[index].x; 760 sum.y+=alpha*filter[i]*source[index].y; 761 sum.z+=alpha*filter[i]*source[index].z; 762 sum.w+=filter[i]*source[index].w; 763 gamma+=alpha*filter[i]; 764 i++; 765 } 766 } 767 768 gamma=1.0/(fabs(gamma) <= MagickEpsilon ? 1.0 : gamma); 769 const ulong index=y*columns+x; 770 destination[index].x=ClampToQuantum(gamma*sum.x); 771 destination[index].y=ClampToQuantum(gamma*sum.y); 772 destination[index].z=ClampToQuantum(gamma*sum.z); 773 destination[index].w=ClampToQuantum(sum.w); 774};</code></pre> 775 776<p>See <a href="https://github.com/ImageMagick/ImageMagick/blob/main/MagickCore/accelerate.c">MagickCore/accelerate.c</a> for a complete implementation of image convolution with an OpenCL kernel.</p> 777 778<p>Note, that under Windows, you might have an issue with TDR (Timeout Detection and Recovery of GPUs). Its purpose is to detect runaway tasks hanging the GPU by using an execution time threshold. For some older low-end GPUs running the OpenCL filters in ImageMagick, longer execution times might trigger the TDR mechanism and pre-empt the GPU image filter. When this happens, ImageMagick automatically falls back to the CPU code path and returns the expected results. To avoid pre-emption, increase the <a href="http://msdn.microsoft.com/en-us/library/windows/hardware/gg487368.aspx">TdrDelay</a> registry key.</p> 779 780<h2><a class="anchor" id="coders"></a>Custom Image Coders</h2> 781 782<p>An image coder (i.e. encoder / decoder) is responsible for registering, optionally classifying, optionally reading, optionally writing, and unregistering one image format (e.g. PNG, GIF, JPEG, etc.). Registering an image coder alerts ImageMagick a particular format is available to read or write. While unregistering tells ImageMagick the format is no longer available. The classifying method looks at the first few bytes of an image and determines if the image is in the expected format. The reader sets the image size, colorspace, and other properties and loads the pixel cache with the pixels. The reader returns a single image or an image sequence (if the format supports multiple images per file), or if an error occurs, an exception and a null image. The writer does the reverse. It takes the image properties and unloads the pixel cache and writes them as required by the image format.</p> 783 784<p>Here is a listing of a sample <a href="https://imagemagick.org/source/mgk.c">custom coder</a>. It reads and writes images in the MGK image format which is simply an ID followed by the image width and height followed by the RGB pixel values.</p> 785<pre class="pre-scrollable highlight"><code>#include <MagickCore/studio.h> 786#include <MagickCore/blob.h> 787#include <MagickCore/cache.h> 788#include <MagickCore/colorspace.h> 789#include <MagickCore/exception.h> 790#include <MagickCore/image.h> 791#include <MagickCore/list.h> 792#include <MagickCore/magick.h> 793#include <MagickCore/memory_.h> 794#include <MagickCore/monitor.h> 795#include <MagickCore/pixel-accessor.h> 796#include <MagickCore/string_.h> 797#include <MagickCore/module.h> 798#include "filter/blob-private.h" 799#include "filter/exception-private.h" 800#include "filter/image-private.h" 801#include "filter/monitor-private.h" 802#include "filter/quantum-private.h" 803 804/* 805 Forward declarations. 806*/ 807static MagickBooleanType 808 WriteMGKImage(const ImageInfo *,Image *,ExceptionInfo *); 809 810/* 811%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% 812% % 813% % 814% % 815% I s M G K % 816% % 817% % 818% % 819%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% 820% 821% IsMGK() returns MagickTrue if the image format type, identified by the 822% magick string, is MGK. 823% 824% The format of the IsMGK method is: 825% 826% MagickBooleanType IsMGK(const unsigned char *magick,const size_t length) 827% 828% A description of each parameter follows: 829% 830% o magick: This string is generally the first few bytes of an image file 831% or blob. 832% 833% o length: Specifies the length of the magick string. 834% 835*/ 836static MagickBooleanType IsMGK(const unsigned char *magick,const size_t length) 837{ 838 if (length < 7) 839 return(MagickFalse); 840 if (LocaleNCompare((char *) magick,"id=mgk",7) == 0) 841 return(MagickTrue); 842 return(MagickFalse); 843} 844 845/* 846%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% 847% % 848% % 849% % 850% R e a d M G K I m a g e % 851% % 852% % 853% % 854%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% 855% 856% ReadMGKImage() reads a MGK image file and returns it. It allocates the 857% memory necessary for the new Image structure and returns a pointer to the 858% new image. 859% 860% The format of the ReadMGKImage method is: 861% 862% Image *ReadMGKImage(const ImageInfo *image_info, 863% ExceptionInfo *exception) 864% 865% A description of each parameter follows: 866% 867% o image_info: the image info. 868% 869% o exception: return any errors or warnings in this structure. 870% 871*/ 872static Image *ReadMGKImage(const ImageInfo *image_info,ExceptionInfo *exception) 873{ 874 char 875 buffer[MaxTextExtent]; 876 877 Image 878 *image; 879 880 long 881 y; 882 883 MagickBooleanType 884 status; 885 886 register long 887 x; 888 889 register Quantum 890 *q; 891 892 register unsigned char 893 *p; 894 895 ssize_t 896 count; 897 898 unsigned char 899 *pixels; 900 901 unsigned long 902 columns, 903 rows; 904 905 /* 906 Open image file. 907 */ 908 assert(image_info != (const ImageInfo *) NULL); 909 assert(image_info->signature == MagickCoreSignature); 910 if (image_info->debug != MagickFalse) 911 (void) LogMagickEvent(TraceEvent,GetMagickModule(),"%s", 912 image_info->filename); 913 assert(exception != (ExceptionInfo *) NULL); 914 assert(exception->signature == MagickCoreSignature); 915 image=AcquireImage(image_info,exception); 916 status=OpenBlob(image_info,image,ReadBinaryBlobMode,exception); 917 if (status == MagickFalse) 918 { 919 image=DestroyImageList(image); 920 return((Image *) NULL); 921 } 922 /* 923 Read MGK image. 924 */ 925 (void) ReadBlobString(image,buffer); /* read magic number */ 926 if (IsMGK(buffer,7) == MagickFalse) 927 ThrowReaderException(CorruptImageError,"ImproperImageHeader"); 928 (void) ReadBlobString(image,buffer); 929 count=(ssize_t) sscanf(buffer,"%lu %lu\n",&columns,&rows); 930 if (count <= 0) 931 ThrowReaderException(CorruptImageError,"ImproperImageHeader"); 932 do 933 { 934 /* 935 Initialize image structure. 936 */ 937 image->columns=columns; 938 image->rows=rows; 939 image->depth=8; 940 if ((image_info->ping != MagickFalse) && (image_info->number_scenes != 0)) 941 if (image->scene >= (image_info->scene+image_info->number_scenes-1)) 942 break; 943 /* 944 Convert MGK raster image to pixel packets. 945 */ 946 if (SetImageExtent(image,image->columns,image->rows,exception) == MagickFalse) 947 return(DestroyImageList(image)); 948 pixels=(unsigned char *) AcquireQuantumMemory((size_t) image->columns, 949 3UL*sizeof(*pixels)); 950 if (pixels == (unsigned char *) NULL) 951 ThrowReaderException(ResourceLimitError,"MemoryAllocationFailed"); 952 for (y=0; y < (long) image->rows; y++) 953 { 954 count=(ssize_t) ReadBlob(image,(size_t) (3*image->columns),pixels); 955 if (count != (ssize_t) (3*image->columns)) 956 ThrowReaderException(CorruptImageError,"UnableToReadImageData"); 957 p=pixels; 958 q=QueueAuthenticPixels(image,0,y,image->columns,1,exception); 959 if (q == (Quantum *) NULL) 960 break; 961 for (x=0; x < (long) image->columns; x++) 962 { 963 SetPixelRed(image,ScaleCharToQuantum(*p++),q); 964 SetPixelGreen(image,ScaleCharToQuantum(*p++),q); 965 SetPixelBlue(image,ScaleCharToQuantum(*p++),q); 966 q+=GetPixelChannels(image); 967 } 968 if (SyncAuthenticPixels(image,exception) == MagickFalse) 969 break; 970 if (image->previous == (Image *) NULL) 971 if ((image->progress_monitor != (MagickProgressMonitor) NULL) && 972 (QuantumTick(y,image->rows) != MagickFalse)) 973 { 974 status=image->progress_monitor(LoadImageTag,y,image->rows, 975 image->client_data); 976 if (status == MagickFalse) 977 break; 978 } 979 } 980 pixels=(unsigned char *) RelinquishMagickMemory(pixels); 981 if (EOFBlob(image) != MagickFalse) 982 { 983 ThrowFileException(exception,CorruptImageError,"UnexpectedEndOfFile", 984 image->filename); 985 break; 986 } 987 /* 988 Proceed to next image. 989 */ 990 if (image_info->number_scenes != 0) 991 if (image->scene >= (image_info->scene+image_info->number_scenes-1)) 992 break; 993 *buffer='\0'; 994 (void) ReadBlobString(image,buffer); 995 count=(ssize_t) sscanf(buffer,"%lu %lu\n",&columns,&rows); 996 if (count > 0) 997 { 998 /* 999 Allocate next image structure. 1000 */ 1001 AcquireNextImage(image_info,image,exception); 1002 if (GetNextImageInList(image) == (Image *) NULL) 1003 { 1004 image=DestroyImageList(image); 1005 return((Image *) NULL); 1006 } 1007 image=SyncNextImageInList(image); 1008 if (image->progress_monitor != (MagickProgressMonitor) NULL) 1009 { 1010 status=SetImageProgress(image,LoadImageTag,TellBlob(image), 1011 GetBlobSize(image)); 1012 if (status == MagickFalse) 1013 break; 1014 } 1015 } 1016 } while (count > 0); 1017 (void) CloseBlob(image); 1018 return(GetFirstImageInList(image)); 1019} 1020 1021/* 1022%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% 1023% % 1024% % 1025% % 1026% R e g i s t e r M G K I m a g e % 1027% % 1028% % 1029% % 1030%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% 1031% 1032% RegisterMGKImage() adds attributes for the MGK image format to 1033% the list of supported formats. The attributes include the image format 1034% tag, a method to read and/or write the format, whether the format 1035% supports the saving of more than one frame to the same file or blob, 1036% whether the format supports native in-memory I/O, and a brief 1037% description of the format. 1038% 1039% The format of the RegisterMGKImage method is: 1040% 1041% unsigned long RegisterMGKImage(void) 1042% 1043*/ 1044ModuleExport unsigned long RegisterMGKImage(void) 1045{ 1046 MagickInfo 1047 *entry; 1048 1049 entry=AcquireMagickInfo("MGK","MGK","MGK image"); 1050 entry->decoder=(DecodeImageHandler *) ReadMGKImage; 1051 entry->encoder=(EncodeImageHandler *) WriteMGKImage; 1052 entry->magick=(IsImageFormatHandler *) IsMGK; 1053 (void) RegisterMagickInfo(entry); 1054 return(MagickImageCoderSignature); 1055} 1056 1057/* 1058%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% 1059% % 1060% % 1061% % 1062% U n r e g i s t e r M G K I m a g e % 1063% % 1064% % 1065% % 1066%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% 1067% 1068% UnregisterMGKImage() removes format registrations made by the 1069% MGK module from the list of supported formats. 1070% 1071% The format of the UnregisterMGKImage method is: 1072% 1073% UnregisterMGKImage(void) 1074% 1075*/ 1076ModuleExport void UnregisterMGKImage(void) 1077{ 1078 (void) UnregisterMagickInfo("MGK"); 1079} 1080 1081/* 1082%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% 1083% % 1084% % 1085% % 1086% W r i t e M G K I m a g e % 1087% % 1088% % 1089% % 1090%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% 1091% 1092% WriteMGKImage() writes an image to a file in red, green, and blue MGK 1093% rasterfile format. 1094% 1095% The format of the WriteMGKImage method is: 1096% 1097% MagickBooleanType WriteMGKImage(const ImageInfo *image_info, 1098% Image *image) 1099% 1100% A description of each parameter follows. 1101% 1102% o image_info: the image info. 1103% 1104% o image: The image. 1105% 1106% o exception: return any errors or warnings in this structure. 1107% 1108*/ 1109static MagickBooleanType WriteMGKImage(const ImageInfo *image_info,Image *image, 1110 ExceptionInfo *exception) 1111{ 1112 char 1113 buffer[MaxTextExtent]; 1114 1115 long 1116 y; 1117 1118 MagickBooleanType 1119 status; 1120 1121 MagickOffsetType 1122 scene; 1123 1124 register const Quantum 1125 *p; 1126 1127 register long 1128 x; 1129 1130 register unsigned char 1131 *q; 1132 1133 unsigned char 1134 *pixels; 1135 1136 /* 1137 Open output image file. 1138 */ 1139 assert(image_info != (const ImageInfo *) NULL); 1140 assert(image_info->signature == MagickCoreSignature); 1141 assert(image != (Image *) NULL); 1142 assert(image->signature == MagickCoreSignature); 1143 if (image->debug != MagickFalse) 1144 (void) LogMagickEvent(TraceEvent,GetMagickModule(),"%s",image->filename); 1145 status=OpenBlob(image_info,image,WriteBinaryBlobMode,exception); 1146 if (status == MagickFalse) 1147 return(status); 1148 scene=0; 1149 do 1150 { 1151 /* 1152 Allocate memory for pixels. 1153 */ 1154 if (image->colorspace != RGBColorspace) 1155 (void) SetImageColorspace(image,RGBColorspace,exception); 1156 pixels=(unsigned char *) AcquireQuantumMemory((size_t) image->columns, 1157 3UL*sizeof(*pixels)); 1158 if (pixels == (unsigned char *) NULL) 1159 ThrowWriterException(ResourceLimitError,"MemoryAllocationFailed"); 1160 /* 1161 Initialize raster file header. 1162 */ 1163 (void) WriteBlobString(image,"id=mgk\n"); 1164 (void) FormatLocaleString(buffer,MaxTextExtent,"%lu %lu\n",image->columns, 1165 image->rows); 1166 (void) WriteBlobString(image,buffer); 1167 for (y=0; y < (long) image->rows; y++) 1168 { 1169 p=GetVirtualPixels(image,0,y,image->columns,1,exception); 1170 if (p == (const Quantum *) NULL) 1171 break; 1172 q=pixels; 1173 for (x=0; x < (long) image->columns; x++) 1174 { 1175 *q++=ScaleQuantumToChar(GetPixelRed(image,p)); 1176 *q++=ScaleQuantumToChar(GetPixelGreen(image,p)); 1177 *q++=ScaleQuantumToChar(GetPixelBlue(image,p)); 1178 p+=GetPixelChannels(image); 1179 } 1180 (void) WriteBlob(image,(size_t) (q-pixels),pixels); 1181 if (image->previous == (Image *) NULL) 1182 if ((image->progress_monitor != (MagickProgressMonitor) NULL) && 1183 (QuantumTick(y,image->rows) != MagickFalse)) 1184 { 1185 status=image->progress_monitor(SaveImageTag,y,image->rows, 1186 image->client_data); 1187 if (status == MagickFalse) 1188 break; 1189 } 1190 } 1191 pixels=(unsigned char *) RelinquishMagickMemory(pixels); 1192 if (GetNextImageInList(image) == (Image *) NULL) 1193 break; 1194 image=SyncNextImageInList(image); 1195 status=SetImageProgress(image,SaveImagesTag,scene, 1196 GetImageListLength(image)); 1197 if (status == MagickFalse) 1198 break; 1199 scene++; 1200 } while (image_info->adjoin != MagickFalse); 1201 (void) CloseBlob(image); 1202 return(MagickTrue); 1203}</code></pre> 1204 1205<p>To invoke the custom coder from the command line, use these commands:</p> 1206<pre class="highlight"><code>magick logo: logo.mgk 1207display logo.mgk 1208</code></pre> 1209 1210<p>We provide the <a href="https://download.imagemagick.org/ImageMagick/download/kits/">Magick Coder Kit</a> to help you get started writing your own custom coder.</p> 1211 1212<h2><a class="anchor" id="filters"></a>Custom Image Filters</h2> 1213 1214<p>ImageMagick provides a convenient mechanism for adding your own custom image processing algorithms. We call these image filters and they are invoked from the command line with the <a href="../www/command-line-options.html#process">-process</a> option or from the MagickCore API method <a href="api/module.html#ExecuteModuleProcess">ExecuteModuleProcess()</a>.</p> 1215 1216<p>Here is a listing of a sample <a href="https://imagemagick.org/source/analyze.c">custom image filter</a>. It computes a few statistics such as the pixel brightness and saturation mean and standard-deviation.</p> 1217<pre class="pre-scrollable highlight"><code>#include <stdio.h> 1218#include <stdlib.h> 1219#include <string.h> 1220#include <time.h> 1221#include <assert.h> 1222#include <math.h> 1223#include "MagickCore/studio.h" 1224#include "MagickCore/MagickCore.h" 1225 1226/* 1227%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% 1228% % 1229% % 1230% % 1231% a n a l y z e I m a g e % 1232% % 1233% % 1234% % 1235%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% 1236% 1237% analyzeImage() computes the brightness and saturation mean, standard 1238% deviation, kurtosis and skewness and stores these values as attributes 1239% of the image. 1240% 1241% The format of the analyzeImage method is: 1242% 1243% size_t analyzeImage(Image *images,const int argc,char **argv, 1244% ExceptionInfo *exception) 1245% 1246% A description of each parameter follows: 1247% 1248% o image: the address of a structure of type Image. 1249% 1250% o argc: Specifies a pointer to an integer describing the number of 1251% elements in the argument vector. 1252% 1253% o argv: Specifies a pointer to a text array containing the command line 1254% arguments. 1255% 1256% o exception: return any errors or warnings in this structure. 1257% 1258*/ 1259 1260typedef struct _StatisticsInfo 1261{ 1262 double 1263 area, 1264 brightness, 1265 mean, 1266 standard_deviation, 1267 sum[5], 1268 kurtosis, 1269 skewness; 1270} StatisticsInfo; 1271 1272static inline int GetMagickNumberThreads(const Image *source, 1273 const Image *destination,const size_t chunk,int multithreaded) 1274{ 1275#define MagickMax(x,y) (((x) > (y)) ? (x) : (y)) 1276#define MagickMin(x,y) (((x) < (y)) ? (x) : (y)) 1277 1278 /* 1279 Number of threads bounded by the amount of work and any thread resource 1280 limit. The limit is 2 if the pixel cache type is not memory or 1281 memory-mapped. 1282 */ 1283 if (multithreaded == 0) 1284 return(1); 1285 if (((GetImagePixelCacheType(source) != MemoryCache) && 1286 (GetImagePixelCacheType(source) != MapCache)) || 1287 ((GetImagePixelCacheType(destination) != MemoryCache) && 1288 (GetImagePixelCacheType(destination) != MapCache))) 1289 return(MagickMax(MagickMin(GetMagickResourceLimit(ThreadResource),2),1)); 1290 return(MagickMax(MagickMin((ssize_t) GetMagickResourceLimit(ThreadResource), 1291 (ssize_t) (chunk)/64),1)); 1292} 1293 1294ModuleExport size_t analyzeImage(Image **images,const int argc, 1295 const char **argv,ExceptionInfo *exception) 1296{ 1297#define AnalyzeImageFilterTag "Filter/Analyze" 1298#define magick_number_threads(source,destination,chunk,multithreaded) \ 1299 num_threads(GetMagickNumberThreads(source,destination,chunk,multithreaded)) 1300 1301 char 1302 text[MagickPathExtent]; 1303 1304 Image 1305 *image; 1306 1307 MagickBooleanType 1308 status; 1309 1310 MagickOffsetType 1311 progress; 1312 1313 assert(images != (Image **) NULL); 1314 assert(*images != (Image *) NULL); 1315 assert((*images)->signature == MagickCoreSignature); 1316 (void) argc; 1317 (void) argv; 1318 image=(*images); 1319 status=MagickTrue; 1320 progress=0; 1321 for ( ; image != (Image *) NULL; image=GetNextImageInList(image)) 1322 { 1323 CacheView 1324 *image_view; 1325 1326 double 1327 area; 1328 1329 ssize_t 1330 y; 1331 1332 StatisticsInfo 1333 brightness, 1334 saturation; 1335 1336 if (status == MagickFalse) 1337 continue; 1338 (void) memset(&brightness,0,sizeof(brightness)); 1339 (void) memset(&saturation,0,sizeof(saturation)); 1340 status=MagickTrue; 1341 image_view=AcquireVirtualCacheView(image,exception); 1342#if defined(MAGICKCORE_OPENMP_SUPPORT) 1343 #pragma omp parallel for schedule(static) \ 1344 shared(progress,status,brightness,saturation) \ 1345 magick_number_threads(image,image,image->rows,1) 1346#endif 1347 for (y=0; y < (ssize_t) image->rows; y++) 1348 { 1349 const Quantum 1350 *p; 1351 1352 ssize_t 1353 i, 1354 x; 1355 1356 StatisticsInfo 1357 local_brightness, 1358 local_saturation; 1359 1360 if (status == MagickFalse) 1361 continue; 1362 p=GetCacheViewVirtualPixels(image_view,0,y,image->columns,1,exception); 1363 if (p == (const Quantum *) NULL) 1364 { 1365 status=MagickFalse; 1366 continue; 1367 } 1368 (void) memset(&local_brightness,0,sizeof(local_brightness)); 1369 (void) memset(&local_saturation,0,sizeof(local_saturation)); 1370 for (x=0; x < (ssize_t) image->columns; x++) 1371 { 1372 double 1373 b, 1374 h, 1375 s; 1376 1377 ConvertRGBToHSL(GetPixelRed(image,p),GetPixelGreen(image,p), 1378 GetPixelBlue(image,p),&h,&s,&b); 1379 b*=QuantumRange; 1380 for (i=1; i <= 4; i++) 1381 local_brightness.sum[i]+=pow(b,(double) i); 1382 s*=QuantumRange; 1383 for (i=1; i <= 4; i++) 1384 local_saturation.sum[i]+=pow(s,(double) i); 1385 p+=GetPixelChannels(image); 1386 } 1387#if defined(MAGICKCORE_OPENMP_SUPPORT) 1388 #pragma omp critical (analyzeImage) 1389#endif 1390 for (i=1; i <= 4; i++) 1391 { 1392 brightness.sum[i]+=local_brightness.sum[i]; 1393 saturation.sum[i]+=local_saturation.sum[i]; 1394 } 1395 } 1396 image_view=DestroyCacheView(image_view); 1397 area=(double) image->columns*image->rows; 1398 brightness.mean=brightness.sum[1]/area; 1399 (void) FormatLocaleString(text,MagickPathExtent,"%g",brightness.mean); 1400 (void) SetImageProperty(image,"filter:brightness:mean",text,exception); 1401 brightness.standard_deviation=sqrt(brightness.sum[2]/area- 1402 (brightness.sum[1]/area*brightness.sum[1]/area)); 1403 (void) FormatLocaleString(text,MagickPathExtent,"%g", 1404 brightness.standard_deviation); 1405 (void) SetImageProperty(image,"filter:brightness:standard-deviation",text, 1406 exception); 1407 if (fabs(brightness.standard_deviation) >= MagickEpsilon) 1408 brightness.kurtosis=(brightness.sum[4]/area-4.0*brightness.mean* 1409 brightness.sum[3]/area+6.0*brightness.mean*brightness.mean* 1410 brightness.sum[2]/area-3.0*brightness.mean*brightness.mean* 1411 brightness.mean*brightness.mean)/(brightness.standard_deviation* 1412 brightness.standard_deviation*brightness.standard_deviation* 1413 brightness.standard_deviation)-3.0; 1414 (void) FormatLocaleString(text,MagickPathExtent,"%g",brightness.kurtosis); 1415 (void) SetImageProperty(image,"filter:brightness:kurtosis",text,exception); 1416 if (brightness.standard_deviation != 0) 1417 brightness.skewness=(brightness.sum[3]/area-3.0*brightness.mean* 1418 brightness.sum[2]/area+2.0*brightness.mean*brightness.mean* 1419 brightness.mean)/(brightness.standard_deviation* 1420 brightness.standard_deviation*brightness.standard_deviation); 1421 (void) FormatLocaleString(text,MagickPathExtent,"%g",brightness.skewness); 1422 (void) SetImageProperty(image,"filter:brightness:skewness",text,exception); 1423 saturation.mean=saturation.sum[1]/area; 1424 (void) FormatLocaleString(text,MagickPathExtent,"%g",saturation.mean); 1425 (void) SetImageProperty(image,"filter:saturation:mean",text,exception); 1426 saturation.standard_deviation=sqrt(saturation.sum[2]/area- 1427 (saturation.sum[1]/area*saturation.sum[1]/area)); 1428 (void) FormatLocaleString(text,MagickPathExtent,"%g", 1429 saturation.standard_deviation); 1430 (void) SetImageProperty(image,"filter:saturation:standard-deviation",text, 1431 exception); 1432 if (fabs(saturation.standard_deviation) >= MagickEpsilon) 1433 saturation.kurtosis=(saturation.sum[4]/area-4.0*saturation.mean* 1434 saturation.sum[3]/area+6.0*saturation.mean*saturation.mean* 1435 saturation.sum[2]/area-3.0*saturation.mean*saturation.mean* 1436 saturation.mean*saturation.mean)/(saturation.standard_deviation* 1437 saturation.standard_deviation*saturation.standard_deviation* 1438 saturation.standard_deviation)-3.0; 1439 (void) FormatLocaleString(text,MagickPathExtent,"%g",saturation.kurtosis); 1440 (void) SetImageProperty(image,"filter:saturation:kurtosis",text,exception); 1441 if (fabs(saturation.standard_deviation) >= MagickEpsilon) 1442 saturation.skewness=(saturation.sum[3]/area-3.0*saturation.mean* 1443 saturation.sum[2]/area+2.0*saturation.mean*saturation.mean* 1444 saturation.mean)/(saturation.standard_deviation* 1445 saturation.standard_deviation*saturation.standard_deviation); 1446 (void) FormatLocaleString(text,MagickPathExtent,"%g",saturation.skewness); 1447 (void) SetImageProperty(image,"filter:saturation:skewness",text,exception); 1448 if (image->progress_monitor != (MagickProgressMonitor) NULL) 1449 { 1450 MagickBooleanType 1451 proceed; 1452 1453#if defined(MAGICKCORE_OPENMP_SUPPORT) 1454 #pragma omp atomic 1455#endif 1456 progress++; 1457 proceed=SetImageProgress(image,AnalyzeImageFilterTag,progress, 1458 GetImageListLength(image)); 1459 if (proceed == MagickFalse) 1460 status=MagickFalse; 1461 } 1462 } 1463 return(MagickImageFilterSignature); 1464}</code></pre> 1465 1466<p>To invoke the custom filter from the command line, use this command:</p> 1467 1468<pre class="highlight"><code>magick logo: -process \"analyze\" -verbose info: 1469 Image: logo: 1470 Format: LOGO (ImageMagick Logo) 1471 Class: PseudoClass 1472 Geometry: 640x480 1473 ... 1474 filter:brightness:kurtosis: 3.97886 1475 filter:brightness:mean: 58901.3 1476 filter:brightness:skewness: -2.30827 1477 filter:brightness:standard-deviation: 16179.8 1478 filter:saturation:kurtosis: 6.59719 1479 filter:saturation:mean: 5321.05 1480 filter:saturation:skewness: 2.75679 1481 filter:saturation:standard-deviation: 14484.7 1482</code></pre> 1483 1484 1485<p>We provide the <a href="https://download.imagemagick.org/ImageMagick/download/kits/">Magick Filter Kit</a> to help you get started writing your own custom image filter.</p> 1486 1487</div> 1488 </div> 1489 </main><!-- /.container --> 1490 <footer class="magick-footer"> 1491 <div class="container-fluid"> 1492 <a href="../www/security-policy.html">Security</a> • 1493 <a href="../www/news.html">News</a> 1494 1495 <a href="architecture.html#"><img class="d-inline" id="wand" alt="And Now a Touch of Magick" width="16" height="16" src="../images/wand.ico"/></a> 1496 1497 <a href="../www/links.html">Related</a> • 1498 <a href="../www/sitemap.html">Sitemap</a> 1499 <br/> 1500 <a href="../www/support.html">Sponsor</a> • 1501 <a href="../www/cite.html">Cite</a> • 1502 <a href="http://pgp.mit.edu/pks/lookup?op=get&search=0x89AB63D48277377A">Public Key</a> • 1503 <a href="../www/https://imagemagick.org/script/contact.php">Contact Us</a> 1504 <br/> 1505 <a href="https://github.com/imagemagick/imagemagick" target="_blank" rel="noopener" aria-label="GitHub"><svg xmlns="http://www.w3.org/2000/svg" class="navbar-nav-svg" viewBox="0 0 512 499.36" width="2%" height="2%" role="img" focusable="false"><title>GitHub</title><path fill="currentColor" fill-rule="evenodd" d="M256 0C114.64 0 0 114.61 0 256c0 113.09 73.34 209 175.08 242.9 12.8 2.35 17.47-5.56 17.47-12.34 0-6.08-.22-22.18-.35-43.54-71.2 15.49-86.2-34.34-86.2-34.34-11.64-29.57-28.42-37.45-28.42-37.45-23.27-15.84 1.73-15.55 1.73-15.55 25.69 1.81 39.21 26.38 39.21 26.38 22.84 39.12 59.92 27.82 74.5 21.27 2.33-16.54 8.94-27.82 16.25-34.22-56.84-6.43-116.6-28.43-116.6-126.49 0-27.95 10-50.8 26.35-68.69-2.63-6.48-11.42-32.5 2.51-67.75 0 0 21.49-6.88 70.4 26.24a242.65 242.65 0 0 1 128.18 0c48.87-33.13 70.33-26.24 70.33-26.24 14 35.25 5.18 61.27 2.55 67.75 16.41 17.9 26.31 40.75 26.31 68.69 0 98.35-59.85 120-116.88 126.32 9.19 7.9 17.38 23.53 17.38 47.41 0 34.22-.31 61.83-.31 70.23 0 6.85 4.61 14.81 17.6 12.31C438.72 464.97 512 369.08 512 256.02 512 114.62 397.37 0 256 0z"/></svg></a> • 1506 <a href="https://twitter.com/imagemagick" target="_blank" rel="noopener" aria-label="Twitter"><svg xmlns="http://www.w3.org/2000/svg" class="navbar-nav-svg" viewBox="0 0 512 416.32" width="2%" height="2%" role="img" focusable="false"><title>Twitter</title><path fill="currentColor" d="M160.83 416.32c193.2 0 298.92-160.22 298.92-298.92 0-4.51 0-9-.2-13.52A214 214 0 0 0 512 49.38a212.93 212.93 0 0 1-60.44 16.6 105.7 105.7 0 0 0 46.3-58.19 209 209 0 0 1-66.79 25.37 105.09 105.09 0 0 0-181.73 71.91 116.12 116.12 0 0 0 2.66 24c-87.28-4.3-164.73-46.3-216.56-109.82A105.48 105.48 0 0 0 68 159.6a106.27 106.27 0 0 1-47.53-13.11v1.43a105.28 105.28 0 0 0 84.21 103.06 105.67 105.67 0 0 1-47.33 1.84 105.06 105.06 0 0 0 98.14 72.94A210.72 210.72 0 0 1 25 370.84a202.17 202.17 0 0 1-25-1.43 298.85 298.85 0 0 0 160.83 46.92"/></svg></a> 1507 <br/> 1508 <small>© 1999-2021 ImageMagick Studio LLC</small> 1509 </div> 1510 </footer> 1511 1512 <!-- Javascript assets --> 1513 <script src="assets/magick.js" ></script> 1514 </body> 1515</html> 1516<!-- Magick Cache 13th February 2021 11:37 -->
