• Home
  • Line#
  • Scopes#
  • Navigate#
  • Raw
  • Download
1 // Ceres Solver - A fast non-linear least squares minimizer
2 // Copyright 2010, 2011, 2012 Google Inc. All rights reserved.
3 // http://code.google.com/p/ceres-solver/
4 //
5 // Redistribution and use in source and binary forms, with or without
6 // modification, are permitted provided that the following conditions are met:
7 //
8 // * Redistributions of source code must retain the above copyright notice,
9 //   this list of conditions and the following disclaimer.
10 // * Redistributions in binary form must reproduce the above copyright notice,
11 //   this list of conditions and the following disclaimer in the documentation
12 //   and/or other materials provided with the distribution.
13 // * Neither the name of Google Inc. nor the names of its contributors may be
14 //   used to endorse or promote products derived from this software without
15 //   specific prior written permission.
16 //
17 // THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
18 // AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
19 // IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
20 // ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
21 // LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
22 // CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
23 // SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
24 // INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
25 // CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
26 // ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
27 // POSSIBILITY OF SUCH DAMAGE.
28 //
29 // Author: sameeragarwal@google.com (Sameer Agarwal)
30 
31 #include <cmath>
32 #include "ceres/fpclassify.h"
33 #include "ceres/internal/autodiff.h"
34 #include "ceres/internal/eigen.h"
35 #include "ceres/local_parameterization.h"
36 #include "ceres/rotation.h"
37 #include "gtest/gtest.h"
38 
39 namespace ceres {
40 namespace internal {
41 
TEST(IdentityParameterization,EverythingTest)42 TEST(IdentityParameterization, EverythingTest) {
43   IdentityParameterization parameterization(3);
44   EXPECT_EQ(parameterization.GlobalSize(), 3);
45   EXPECT_EQ(parameterization.LocalSize(), 3);
46 
47   double x[3] = {1.0, 2.0, 3.0};
48   double delta[3] = {0.0, 1.0, 2.0};
49   double x_plus_delta[3] = {0.0, 0.0, 0.0};
50   parameterization.Plus(x, delta, x_plus_delta);
51   EXPECT_EQ(x_plus_delta[0], 1.0);
52   EXPECT_EQ(x_plus_delta[1], 3.0);
53   EXPECT_EQ(x_plus_delta[2], 5.0);
54 
55   double jacobian[9];
56   parameterization.ComputeJacobian(x, jacobian);
57   int k = 0;
58   for (int i = 0; i < 3; ++i) {
59     for (int j = 0; j < 3; ++j, ++k) {
60       EXPECT_EQ(jacobian[k], (i == j) ? 1.0 : 0.0);
61     }
62   }
63 }
64 
TEST(SubsetParameterization,DeathTests)65 TEST(SubsetParameterization, DeathTests) {
66   vector<int> constant_parameters;
67   EXPECT_DEATH_IF_SUPPORTED(
68       SubsetParameterization parameterization(1, constant_parameters),
69       "at least");
70 
71   constant_parameters.push_back(0);
72   EXPECT_DEATH_IF_SUPPORTED(
73       SubsetParameterization parameterization(1, constant_parameters),
74       "Number of parameters");
75 
76   constant_parameters.push_back(1);
77   EXPECT_DEATH_IF_SUPPORTED(
78       SubsetParameterization parameterization(2, constant_parameters),
79       "Number of parameters");
80 
81   constant_parameters.push_back(1);
82   EXPECT_DEATH_IF_SUPPORTED(
83       SubsetParameterization parameterization(2, constant_parameters),
84       "duplicates");
85 }
86 
TEST(SubsetParameterization,NormalFunctionTest)87 TEST(SubsetParameterization, NormalFunctionTest) {
88   double x[4] = {1.0, 2.0, 3.0, 4.0};
89   for (int i = 0; i < 4; ++i) {
90     vector<int> constant_parameters;
91     constant_parameters.push_back(i);
92     SubsetParameterization parameterization(4, constant_parameters);
93     double delta[3] = {1.0, 2.0, 3.0};
94     double x_plus_delta[4] = {0.0, 0.0, 0.0};
95 
96     parameterization.Plus(x, delta, x_plus_delta);
97     int k = 0;
98     for (int j = 0; j < 4; ++j) {
99       if (j == i)  {
100         EXPECT_EQ(x_plus_delta[j], x[j]);
101       } else {
102         EXPECT_EQ(x_plus_delta[j], x[j] + delta[k++]);
103       }
104     }
105 
106     double jacobian[4 * 3];
107     parameterization.ComputeJacobian(x, jacobian);
108     int delta_cursor = 0;
109     int jacobian_cursor = 0;
110     for (int j = 0; j < 4; ++j) {
111       if (j != i) {
112         for (int k = 0; k < 3; ++k, jacobian_cursor++) {
113           EXPECT_EQ(jacobian[jacobian_cursor], delta_cursor == k ? 1.0 : 0.0);
114         }
115         ++delta_cursor;
116       } else {
117         for (int k = 0; k < 3; ++k, jacobian_cursor++) {
118           EXPECT_EQ(jacobian[jacobian_cursor], 0.0);
119         }
120       }
121     }
122   };
123 }
124 
125 // Functor needed to implement automatically differentiated Plus for
126 // quaternions.
127 struct QuaternionPlus {
128   template<typename T>
operator ()ceres::internal::QuaternionPlus129   bool operator()(const T* x, const T* delta, T* x_plus_delta) const {
130     const T squared_norm_delta =
131         delta[0] * delta[0] + delta[1] * delta[1] + delta[2] * delta[2];
132 
133     T q_delta[4];
134     if (squared_norm_delta > T(0.0)) {
135       T norm_delta = sqrt(squared_norm_delta);
136       const T sin_delta_by_delta = sin(norm_delta) / norm_delta;
137       q_delta[0] = cos(norm_delta);
138       q_delta[1] = sin_delta_by_delta * delta[0];
139       q_delta[2] = sin_delta_by_delta * delta[1];
140       q_delta[3] = sin_delta_by_delta * delta[2];
141     } else {
142       // We do not just use q_delta = [1,0,0,0] here because that is a
143       // constant and when used for automatic differentiation will
144       // lead to a zero derivative. Instead we take a first order
145       // approximation and evaluate it at zero.
146       q_delta[0] = T(1.0);
147       q_delta[1] = delta[0];
148       q_delta[2] = delta[1];
149       q_delta[3] = delta[2];
150     }
151 
152     QuaternionProduct(q_delta, x, x_plus_delta);
153     return true;
154   }
155 };
156 
QuaternionParameterizationTestHelper(const double * x,const double * delta,const double * q_delta)157 void QuaternionParameterizationTestHelper(const double* x,
158                                           const double* delta,
159                                           const double* q_delta) {
160   const double kTolerance = 1e-14;
161   double x_plus_delta_ref[4] = {0.0, 0.0, 0.0, 0.0};
162   QuaternionProduct(q_delta, x, x_plus_delta_ref);
163 
164   double x_plus_delta[4] = {0.0, 0.0, 0.0, 0.0};
165   QuaternionParameterization param;
166   param.Plus(x, delta, x_plus_delta);
167   for (int i = 0; i < 4; ++i) {
168     EXPECT_NEAR(x_plus_delta[i], x_plus_delta_ref[i], kTolerance);
169   }
170 
171   const double x_plus_delta_norm =
172       sqrt(x_plus_delta[0] * x_plus_delta[0] +
173            x_plus_delta[1] * x_plus_delta[1] +
174            x_plus_delta[2] * x_plus_delta[2] +
175            x_plus_delta[3] * x_plus_delta[3]);
176 
177   EXPECT_NEAR(x_plus_delta_norm, 1.0, kTolerance);
178 
179   double jacobian_ref[12];
180   double zero_delta[3] = {0.0, 0.0, 0.0};
181   const double* parameters[2] = {x, zero_delta};
182   double* jacobian_array[2] = { NULL, jacobian_ref };
183 
184   // Autodiff jacobian at delta_x = 0.
185   internal::AutoDiff<QuaternionPlus, double, 4, 3>::Differentiate(
186       QuaternionPlus(), parameters, 4, x_plus_delta, jacobian_array);
187 
188   double jacobian[12];
189   param.ComputeJacobian(x, jacobian);
190   for (int i = 0; i < 12; ++i) {
191     EXPECT_TRUE(IsFinite(jacobian[i]));
192     EXPECT_NEAR(jacobian[i], jacobian_ref[i], kTolerance)
193         << "Jacobian mismatch: i = " << i
194         << "\n Expected \n" << ConstMatrixRef(jacobian_ref, 4, 3)
195         << "\n Actual \n" << ConstMatrixRef(jacobian, 4, 3);
196   }
197 }
198 
TEST(QuaternionParameterization,ZeroTest)199 TEST(QuaternionParameterization, ZeroTest) {
200   double x[4] = {0.5, 0.5, 0.5, 0.5};
201   double delta[3] = {0.0, 0.0, 0.0};
202   double q_delta[4] = {1.0, 0.0, 0.0, 0.0};
203   QuaternionParameterizationTestHelper(x, delta, q_delta);
204 }
205 
206 
TEST(QuaternionParameterization,NearZeroTest)207 TEST(QuaternionParameterization, NearZeroTest) {
208   double x[4] = {0.52, 0.25, 0.15, 0.45};
209   double norm_x = sqrt(x[0] * x[0] +
210                        x[1] * x[1] +
211                        x[2] * x[2] +
212                        x[3] * x[3]);
213   for (int i = 0; i < 4; ++i) {
214     x[i] = x[i] / norm_x;
215   }
216 
217   double delta[3] = {0.24, 0.15, 0.10};
218   for (int i = 0; i < 3; ++i) {
219     delta[i] = delta[i] * 1e-14;
220   }
221 
222   double q_delta[4];
223   q_delta[0] = 1.0;
224   q_delta[1] = delta[0];
225   q_delta[2] = delta[1];
226   q_delta[3] = delta[2];
227 
228   QuaternionParameterizationTestHelper(x, delta, q_delta);
229 }
230 
TEST(QuaternionParameterization,AwayFromZeroTest)231 TEST(QuaternionParameterization, AwayFromZeroTest) {
232   double x[4] = {0.52, 0.25, 0.15, 0.45};
233   double norm_x = sqrt(x[0] * x[0] +
234                        x[1] * x[1] +
235                        x[2] * x[2] +
236                        x[3] * x[3]);
237 
238   for (int i = 0; i < 4; ++i) {
239     x[i] = x[i] / norm_x;
240   }
241 
242   double delta[3] = {0.24, 0.15, 0.10};
243   const double delta_norm = sqrt(delta[0] * delta[0] +
244                                  delta[1] * delta[1] +
245                                  delta[2] * delta[2]);
246   double q_delta[4];
247   q_delta[0] = cos(delta_norm);
248   q_delta[1] = sin(delta_norm) / delta_norm * delta[0];
249   q_delta[2] = sin(delta_norm) / delta_norm * delta[1];
250   q_delta[3] = sin(delta_norm) / delta_norm * delta[2];
251 
252   QuaternionParameterizationTestHelper(x, delta, q_delta);
253 }
254 
255 
256 }  // namespace internal
257 }  // namespace ceres
258