1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
/*M///////////////////////////////////////////////////////////////////////////////////////
//
// IMPORTANT: READ BEFORE DOWNLOADING, COPYING, INSTALLING OR USING.
//
// By downloading, copying, installing or using the software you agree to this license.
// If you do not agree to this license, do not download, install,
// copy or use the software.
//
//
// Intel License Agreement
// For Open Source Computer Vision Library
//
// Copyright (C) 2000, Intel Corporation, all rights reserved.
// Third party copyrights are property of their respective owners.
//
// Redistribution and use in source and binary forms, with or without modification,
// are permitted provided that the following conditions are met:
//
// * Redistribution's of source code must retain the above copyright notice,
// this list of conditions and the following disclaimer.
//
// * Redistribution's in binary form must reproduce the above copyright notice,
// this list of conditions and the following disclaimer in the documentation
// and/or other materials provided with the distribution.
//
// * The name of Intel Corporation may not be used to endorse or promote products
// derived from this software without specific prior written permission.
//
// This software is provided by the copyright holders and contributors "as is" and
// any express or implied warranties, including, but not limited to, the implied
// warranties of merchantability and fitness for a particular purpose are disclaimed.
// In no event shall the Intel Corporation or contributors be liable for any direct,
// indirect, incidental, special, exemplary, or consequential damages
// (including, but not limited to, procurement of substitute goods or services;
// loss of use, data, or profits; or business interruption) however caused
// and on any theory of liability, whether in contract, strict liability,
// or tort (including negligence or otherwise) arising in any way out of
// the use of this software, even if advised of the possibility of such damage.
//
//M*/
#include "precomp.hpp"
#include <limits>
#include <utility>
#include <algorithm>
#include <math.h>
//#define _SUBPIX_VERBOSE
#undef max
namespace cv {
// static void drawCircles(Mat& img, const vector<Point2f>& corners, const vector<float>& radius)
// {
// for(size_t i = 0; i < corners.size(); i++)
// {
// circle(img, corners[i], cvRound(radius[i]), CV_RGB(255, 0, 0));
// }
// }
// static int histQuantile(const Mat& hist, float quantile)
// {
// if(hist.dims > 1) return -1; // works for 1D histograms only
// float cur_sum = 0;
// float total_sum = (float)sum(hist).val[0];
// float quantile_sum = total_sum*quantile;
// for(int j = 0; j < hist.size[0]; j++)
// {
// cur_sum += (float)hist.at<float>(j);
// if(cur_sum > quantile_sum)
// {
// return j;
// }
// }
// return hist.size[0] - 1;
// }
inline bool is_smaller(const std::pair<int, float>& p1, const std::pair<int, float>& p2)
{
return p1.second < p2.second;
}
static void orderContours(const vector<vector<Point> >& contours, Point2f point, vector<std::pair<int, float> >& order)
{
order.clear();
size_t i, j, n = contours.size();
for(i = 0; i < n; i++)
{
size_t ni = contours[i].size();
double min_dist = std::numeric_limits<double>::max();
for(j = 0; j < ni; j++)
{
double dist = norm(Point2f((float)contours[i][j].x, (float)contours[i][j].y) - point);
min_dist = MIN(min_dist, dist);
}
order.push_back(std::pair<int, float>((int)i, (float)min_dist));
}
std::sort(order.begin(), order.end(), is_smaller);
}
// fit second order curve to a set of 2D points
inline void fitCurve2Order(const vector<Point2f>& /*points*/, vector<float>& /*curve*/)
{
// TBD
}
inline void findCurvesCross(const vector<float>& /*curve1*/, const vector<float>& /*curve2*/, Point2f& /*cross_point*/)
{
}
static void findLinesCrossPoint(Point2f origin1, Point2f dir1, Point2f origin2, Point2f dir2, Point2f& cross_point)
{
float det = dir2.x*dir1.y - dir2.y*dir1.x;
Point2f offset = origin2 - origin1;
float alpha = (dir2.x*offset.y - dir2.y*offset.x)/det;
cross_point = origin1 + dir1*alpha;
}
// static void findCorner(const vector<Point>& contour, Point2f point, Point2f& corner)
// {
// // find the nearest point
// double min_dist = std::numeric_limits<double>::max();
// int min_idx = -1;
// // find corner idx
// for(size_t i = 0; i < contour.size(); i++)
// {
// double dist = norm(Point2f((float)contour[i].x, (float)contour[i].y) - point);
// if(dist < min_dist)
// {
// min_dist = dist;
// min_idx = (int)i;
// }
// }
// assert(min_idx >= 0);
// // temporary solution, have to make something more precise
// corner = contour[min_idx];
// return;
// }
static void findCorner(const vector<Point2f>& contour, Point2f point, Point2f& corner)
{
// find the nearest point
double min_dist = std::numeric_limits<double>::max();
int min_idx = -1;
// find corner idx
for(size_t i = 0; i < contour.size(); i++)
{
double dist = norm(contour[i] - point);
if(dist < min_dist)
{
min_dist = dist;
min_idx = (int)i;
}
}
assert(min_idx >= 0);
// temporary solution, have to make something more precise
corner = contour[min_idx];
return;
}
static int segment_hist_max(const Mat& hist, int& low_thresh, int& high_thresh)
{
Mat bw;
//const double max_bell_width = 20; // we expect two bells with width bounded above
//const double min_bell_width = 5; // and below
double total_sum = sum(hist).val[0];
//double thresh = total_sum/(2*max_bell_width)*0.25f; // quarter of a bar inside a bell
// threshold(hist, bw, thresh, 255.0, CV_THRESH_BINARY);
double quantile_sum = 0.0;
//double min_quantile = 0.2;
double low_sum = 0;
double max_segment_length = 0;
int max_start_x = -1;
int max_end_x = -1;
int start_x = 0;
const double out_of_bells_fraction = 0.1;
for(int x = 0; x < hist.size[0]; x++)
{
quantile_sum += hist.at<float>(x);
if(quantile_sum < 0.2*total_sum) continue;
if(quantile_sum - low_sum > out_of_bells_fraction*total_sum)
{
if(max_segment_length < x - start_x)
{
max_segment_length = x - start_x;
max_start_x = start_x;
max_end_x = x;
}
low_sum = quantile_sum;
start_x = x;
}
}
if(start_x == -1)
{
return 0;
}
else
{
low_thresh = cvRound(max_start_x + 0.25*(max_end_x - max_start_x));
high_thresh = cvRound(max_start_x + 0.75*(max_end_x - max_start_x));
return 1;
}
}
}
bool cv::find4QuadCornerSubpix(InputArray _img, InputOutputArray _corners, Size region_size)
{
Mat img = _img.getMat(), cornersM = _corners.getMat();
int ncorners = cornersM.checkVector(2, CV_32F);
CV_Assert( ncorners >= 0 );
Point2f* corners = cornersM.ptr<Point2f>();
const int nbins = 256;
float ranges[] = {0, 256};
const float* _ranges = ranges;
Mat hist;
#if defined(_SUBPIX_VERBOSE)
vector<float> radius;
radius.assign(corners.size(), 0.0f);
#endif //_SUBPIX_VERBOSE
Mat black_comp, white_comp;
for(int i = 0; i < ncorners; i++)
{
int channels = 0;
Rect roi(cvRound(corners[i].x - region_size.width), cvRound(corners[i].y - region_size.height),
region_size.width*2 + 1, region_size.height*2 + 1);
Mat img_roi = img(roi);
calcHist(&img_roi, 1, &channels, Mat(), hist, 1, &nbins, &_ranges);
#if 0
int black_thresh = histQuantile(hist, 0.45f);
int white_thresh = histQuantile(hist, 0.55f);
#else
int black_thresh = 0, white_thresh = 0;
segment_hist_max(hist, black_thresh, white_thresh);
#endif
threshold(img, black_comp, black_thresh, 255.0, CV_THRESH_BINARY_INV);
threshold(img, white_comp, white_thresh, 255.0, CV_THRESH_BINARY);
const int erode_count = 1;
erode(black_comp, black_comp, Mat(), Point(-1, -1), erode_count);
erode(white_comp, white_comp, Mat(), Point(-1, -1), erode_count);
#if defined(_SUBPIX_VERBOSE)
namedWindow("roi", 1);
imshow("roi", img_roi);
imwrite("test.jpg", img);
namedWindow("black", 1);
imshow("black", black_comp);
namedWindow("white", 1);
imshow("white", white_comp);
cvWaitKey(0);
imwrite("black.jpg", black_comp);
imwrite("white.jpg", white_comp);
#endif
vector<vector<Point> > white_contours, black_contours;
vector<Vec4i> white_hierarchy, black_hierarchy;
findContours(black_comp, black_contours, black_hierarchy, CV_RETR_LIST, CV_CHAIN_APPROX_SIMPLE);
findContours(white_comp, white_contours, white_hierarchy, CV_RETR_LIST, CV_CHAIN_APPROX_SIMPLE);
if(black_contours.size() < 5 || white_contours.size() < 5) continue;
// find two white and black blobs that are close to the input point
vector<std::pair<int, float> > white_order, black_order;
orderContours(black_contours, corners[i], black_order);
orderContours(white_contours, corners[i], white_order);
const float max_dist = 10.0f;
if(black_order[0].second > max_dist || black_order[1].second > max_dist ||
white_order[0].second > max_dist || white_order[1].second > max_dist)
{
continue; // there will be no improvement in this corner position
}
const vector<Point>* quads[4] = {&black_contours[black_order[0].first], &black_contours[black_order[1].first],
&white_contours[white_order[0].first], &white_contours[white_order[1].first]};
vector<Point2f> quads_approx[4];
Point2f quad_corners[4];
for(int k = 0; k < 4; k++)
{
#if 1
vector<Point2f> temp;
for(size_t j = 0; j < quads[k]->size(); j++) temp.push_back((*quads[k])[j]);
approxPolyDP(Mat(temp), quads_approx[k], 0.5, true);
findCorner(quads_approx[k], corners[i], quad_corners[k]);
#else
findCorner(*quads[k], corners[i], quad_corners[k]);
#endif
quad_corners[k] += Point2f(0.5f, 0.5f);
}
// cross two lines
Point2f origin1 = quad_corners[0];
Point2f dir1 = quad_corners[1] - quad_corners[0];
Point2f origin2 = quad_corners[2];
Point2f dir2 = quad_corners[3] - quad_corners[2];
double angle = acos(dir1.dot(dir2)/(norm(dir1)*norm(dir2)));
if(cvIsNaN(angle) || cvIsInf(angle) || angle < 0.5 || angle > CV_PI - 0.5) continue;
findLinesCrossPoint(origin1, dir1, origin2, dir2, corners[i]);
#if defined(_SUBPIX_VERBOSE)
radius[i] = norm(corners[i] - ground_truth_corners[ground_truth_idx])*6;
#if 1
Mat test(img.size(), CV_32FC3);
cvtColor(img, test, CV_GRAY2RGB);
// line(test, quad_corners[0] - corners[i] + Point2f(30, 30), quad_corners[1] - corners[i] + Point2f(30, 30), cvScalar(0, 255, 0));
// line(test, quad_corners[2] - corners[i] + Point2f(30, 30), quad_corners[3] - corners[i] + Point2f(30, 30), cvScalar(0, 255, 0));
vector<vector<Point> > contrs;
contrs.resize(1);
for(int k = 0; k < 4; k++)
{
//contrs[0] = quads_approx[k];
contrs[0].clear();
for(size_t j = 0; j < quads_approx[k].size(); j++) contrs[0].push_back(quads_approx[k][j]);
drawContours(test, contrs, 0, CV_RGB(0, 0, 255), 1, 1, vector<Vec4i>(), 2);
circle(test, quad_corners[k], 0.5, CV_RGB(255, 0, 0));
}
Mat test1 = test(Rect(corners[i].x - 30, corners[i].y - 30, 60, 60));
namedWindow("1", 1);
imshow("1", test1);
imwrite("test.jpg", test);
waitKey(0);
#endif
#endif //_SUBPIX_VERBOSE
}
#if defined(_SUBPIX_VERBOSE)
Mat test(img.size(), CV_32FC3);
cvtColor(img, test, CV_GRAY2RGB);
drawCircles(test, corners, radius);
namedWindow("corners", 1);
imshow("corners", test);
waitKey();
#endif //_SUBPIX_VERBOSE
return true;
}