1 /*
2 * Copyright (c) 2015 Paul B Mahol
3 *
4 * This file is part of FFmpeg.
5 *
6 * FFmpeg is free software; you can redistribute it and/or
7 * modify it under the terms of the GNU Lesser General Public
8 * License as published by the Free Software Foundation; either
9 * version 2.1 of the License, or (at your option) any later version.
10 *
11 * FFmpeg is distributed in the hope that it will be useful,
12 * but WITHOUT ANY WARRANTY; without even the implied warranty of
13 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
14 * Lesser General Public License for more details.
15 *
16 * You should have received a copy of the GNU Lesser General Public
17 * License along with FFmpeg; if not, write to the Free Software
18 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
19 */
20
21 #include <tesseract/capi.h>
22
23 #include "libavutil/opt.h"
24 #include "avfilter.h"
25 #include "formats.h"
26 #include "internal.h"
27 #include "video.h"
28
29 typedef struct OCRContext {
30 const AVClass *class;
31
32 char *datapath;
33 char *language;
34 char *whitelist;
35 char *blacklist;
36
37 TessBaseAPI *tess;
38 } OCRContext;
39
40 #define OFFSET(x) offsetof(OCRContext, x)
41 #define FLAGS AV_OPT_FLAG_VIDEO_PARAM|AV_OPT_FLAG_FILTERING_PARAM
42
43 static const AVOption ocr_options[] = {
44 { "datapath", "set datapath", OFFSET(datapath), AV_OPT_TYPE_STRING, {.str=NULL}, 0, 0, FLAGS },
45 { "language", "set language", OFFSET(language), AV_OPT_TYPE_STRING, {.str="eng"}, 0, 0, FLAGS },
46 { "whitelist", "set character whitelist", OFFSET(whitelist), AV_OPT_TYPE_STRING, {.str="0123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ.:;,-+_!?\"'[]{}()<>|/\\=*&%$#@!~ "}, 0, 0, FLAGS },
47 { "blacklist", "set character blacklist", OFFSET(blacklist), AV_OPT_TYPE_STRING, {.str=""}, 0, 0, FLAGS },
48 { NULL }
49 };
50
init(AVFilterContext * ctx)51 static av_cold int init(AVFilterContext *ctx)
52 {
53 OCRContext *s = ctx->priv;
54
55 s->tess = TessBaseAPICreate();
56 if (TessBaseAPIInit3(s->tess, s->datapath, s->language) == -1) {
57 av_log(ctx, AV_LOG_ERROR, "failed to init tesseract\n");
58 return AVERROR(EINVAL);
59 }
60
61 if (!TessBaseAPISetVariable(s->tess, "tessedit_char_whitelist", s->whitelist)) {
62 av_log(ctx, AV_LOG_ERROR, "failed to set whitelist\n");
63 return AVERROR(EINVAL);
64 }
65
66 if (!TessBaseAPISetVariable(s->tess, "tessedit_char_blacklist", s->blacklist)) {
67 av_log(ctx, AV_LOG_ERROR, "failed to set blacklist\n");
68 return AVERROR(EINVAL);
69 }
70
71 av_log(ctx, AV_LOG_DEBUG, "Tesseract version: %s\n", TessVersion());
72
73 return 0;
74 }
75
query_formats(AVFilterContext * ctx)76 static int query_formats(AVFilterContext *ctx)
77 {
78 static const enum AVPixelFormat pix_fmts[] = {
79 AV_PIX_FMT_GRAY8,
80 AV_PIX_FMT_YUV410P, AV_PIX_FMT_YUV411P,
81 AV_PIX_FMT_YUV420P, AV_PIX_FMT_YUV422P,
82 AV_PIX_FMT_YUV440P, AV_PIX_FMT_YUV444P,
83 AV_PIX_FMT_YUVJ420P, AV_PIX_FMT_YUVJ422P,
84 AV_PIX_FMT_YUVJ440P, AV_PIX_FMT_YUVJ444P,
85 AV_PIX_FMT_YUVJ411P,
86 AV_PIX_FMT_YUVA444P, AV_PIX_FMT_YUVA422P, AV_PIX_FMT_YUVA420P,
87 AV_PIX_FMT_NONE
88 };
89
90 AVFilterFormats *fmts_list = ff_make_format_list(pix_fmts);
91 if (!fmts_list)
92 return AVERROR(ENOMEM);
93 return ff_set_common_formats(ctx, fmts_list);
94 }
95
filter_frame(AVFilterLink * inlink,AVFrame * in)96 static int filter_frame(AVFilterLink *inlink, AVFrame *in)
97 {
98 AVDictionary **metadata = &in->metadata;
99 AVFilterContext *ctx = inlink->dst;
100 AVFilterLink *outlink = ctx->outputs[0];
101 OCRContext *s = ctx->priv;
102 char *result;
103 int *confs;
104
105 result = TessBaseAPIRect(s->tess, in->data[0], 1,
106 in->linesize[0], 0, 0, in->width, in->height);
107 confs = TessBaseAPIAllWordConfidences(s->tess);
108 av_dict_set(metadata, "lavfi.ocr.text", result, 0);
109 for (int i = 0; confs[i] != -1; i++) {
110 char number[256];
111
112 snprintf(number, sizeof(number), "%d ", confs[i]);
113 av_dict_set(metadata, "lavfi.ocr.confidence", number, AV_DICT_APPEND);
114 }
115
116 TessDeleteText(result);
117 TessDeleteIntArray(confs);
118
119 return ff_filter_frame(outlink, in);
120 }
121
uninit(AVFilterContext * ctx)122 static av_cold void uninit(AVFilterContext *ctx)
123 {
124 OCRContext *s = ctx->priv;
125
126 TessBaseAPIEnd(s->tess);
127 TessBaseAPIDelete(s->tess);
128 }
129
130 AVFILTER_DEFINE_CLASS(ocr);
131
132 static const AVFilterPad ocr_inputs[] = {
133 {
134 .name = "default",
135 .type = AVMEDIA_TYPE_VIDEO,
136 .filter_frame = filter_frame,
137 },
138 { NULL }
139 };
140
141 static const AVFilterPad ocr_outputs[] = {
142 {
143 .name = "default",
144 .type = AVMEDIA_TYPE_VIDEO,
145 },
146 { NULL }
147 };
148
149 AVFilter ff_vf_ocr = {
150 .name = "ocr",
151 .description = NULL_IF_CONFIG_SMALL("Optical Character Recognition."),
152 .priv_size = sizeof(OCRContext),
153 .priv_class = &ocr_class,
154 .query_formats = query_formats,
155 .init = init,
156 .uninit = uninit,
157 .inputs = ocr_inputs,
158 .outputs = ocr_outputs,
159 };
160