-
Notifications
You must be signed in to change notification settings - Fork 5
/
CamLayoutAnalyzer.java
255 lines (230 loc) · 11.3 KB
/
CamLayoutAnalyzer.java
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
package org.genericsystem.layout;
import java.util.ArrayList;
import java.util.Collections;
import java.util.List;
import java.util.concurrent.Executors;
import java.util.concurrent.ScheduledExecutorService;
import java.util.concurrent.TimeUnit;
import java.util.function.Predicate;
import java.util.stream.Collectors;
import org.genericsystem.cv.AbstractApp;
import org.genericsystem.cv.Img;
import org.genericsystem.cv.Tools;
import org.opencv.calib3d.Calib3d;
import org.opencv.core.Core;
import org.opencv.core.CvType;
import org.opencv.core.DMatch;
import org.opencv.core.KeyPoint;
import org.opencv.core.Mat;
import org.opencv.core.MatOfDMatch;
import org.opencv.core.MatOfKeyPoint;
import org.opencv.core.MatOfPoint;
import org.opencv.core.MatOfPoint2f;
import org.opencv.core.Point;
import org.opencv.core.Rect;
import org.opencv.core.RotatedRect;
import org.opencv.core.Scalar;
import org.opencv.core.Size;
import org.opencv.features2d.DescriptorExtractor;
import org.opencv.features2d.DescriptorMatcher;
import org.opencv.imgproc.Imgproc;
import org.opencv.utils.Converters;
import org.opencv.videoio.VideoCapture;
import javafx.scene.image.ImageView;
import javafx.scene.layout.GridPane;
public class CamLayoutAnalyzer extends AbstractApp {
static {
System.loadLibrary(Core.NATIVE_LIBRARY_NAME);
}
private double crop = 0.10;
public static void main(String[] args) {
launch(args);
}
private final VideoCapture capture = new VideoCapture(0);
private final ScheduledExecutorService timer = Executors.newSingleThreadScheduledExecutor();
@Override
protected void fillGrid(GridPane mainGrid) {
// FeatureDetector detector = FeatureDetector.create(FeatureDetector.FAST);
DescriptorExtractor extractor = DescriptorExtractor.create(DescriptorExtractor.ORB);
DescriptorMatcher matcher = DescriptorMatcher.create(DescriptorMatcher.BRUTEFORCE_HAMMING);
Mat frame = new Mat();
capture.read(frame);
ImageView src0 = new ImageView(Tools.mat2jfxImage(frame));
ImageView src1 = new ImageView(Tools.mat2jfxImage(frame));
Img deskewed = deskew(frame, new double[1]);
ImageView src2 = new ImageView(deskewed.toJfxImage());
ImageView src3 = new ImageView(deskewed.toJfxImage());
mainGrid.add(src0, 0, 0);
mainGrid.add(src1, 0, 1);
mainGrid.add(src2, 1, 0);
mainGrid.add(src3, 1, 1);
MatOfKeyPoint[] oldKeypoints = new MatOfKeyPoint[] { detect(deskewed.getSrc()) };
Mat[] oldDescriptors = new Mat[] { new Mat() };
extractor.compute(deskewed.getSrc(), oldKeypoints[0], oldDescriptors[0]);
int[] count = new int[] { 0 };
Mat stabilizedMat = new Mat();
Layout[] layout = new Layout[] { null };
timer.scheduleAtFixedRate(() -> {
MatOfKeyPoint newKeypoints = null;
Mat newDescriptors = null;
try {
capture.read(frame);
double[] angle = new double[1];
Size newSize = new Size(frame.width() * (1 - 2 * crop), frame.height() * (1 - 2 * crop));
Img frameImg = new Img(frame, false).resize(newSize);
src0.setImage(frameImg.toJfxImage());
Img deskewed_ = deskew(frame, angle);
Img deskiewedCopy = new Img(deskewed_.getSrc(), true);
src1.setImage(deskewed_.toJfxImage());
newKeypoints = detect(deskewed_.getSrc());
newDescriptors = new Mat();
extractor.compute(deskewed_.getSrc(), newKeypoints, newDescriptors);
deskewed_.buildLayout().draw(deskiewedCopy, new Scalar(0, 255, 0), 1);
src2.setImage(Tools.mat2jfxImage(deskiewedCopy.getSrc()));
Img stabilized = stabilize(frame, stabilizedMat, deskewed_.size(), matcher, oldKeypoints[0], newKeypoints, oldDescriptors[0], newDescriptors, angle[0], crop);
if (stabilized != null) {
Img stabilizedCopy = new Img(stabilized.getSrc(), true);
if (layout[0] == null)
layout[0] = stabilized.buildLayout();
layout[0].draw(stabilizedCopy, new Scalar(0, 255, 0), 1);
src3.setImage(stabilizedCopy.toJfxImage());
}
count[0]++;
} catch (Exception e) {
e.printStackTrace();
} finally {
if ((count[0] % 50) == 0) {
oldKeypoints[0] = newKeypoints;
oldDescriptors[0] = newDescriptors;
layout[0] = null;
}
}
}, 0, 33, TimeUnit.MILLISECONDS);
}
private Img stabilize(Mat frame, Mat stabilized, Size size, DescriptorMatcher matcher, MatOfKeyPoint oldKeypoints, MatOfKeyPoint newKeypoints, Mat oldDescriptors, Mat newDescriptors, double angle, double crop) {
MatOfDMatch matches = new MatOfDMatch();
matcher.match(oldDescriptors, newDescriptors, matches);
List<DMatch> goodMatches = new ArrayList<>();
for (DMatch dMatch : matches.toArray()) {
if (dMatch.distance <= 40) {
goodMatches.add(dMatch);
}
}
List<KeyPoint> newKeypoints_ = newKeypoints.toList();
List<KeyPoint> oldKeypoints_ = oldKeypoints.toList();
// System.out.println(goodMatches.size() + " " + newKeypoints_.size() + " " + oldKeypoints_.size());
List<Point> goodNewKeypoints = new ArrayList<>();
List<Point> goodOldKeypoints = new ArrayList<>();
for (DMatch goodMatch : goodMatches) {
goodNewKeypoints.add(newKeypoints_.get(goodMatch.trainIdx).pt);
goodOldKeypoints.add(oldKeypoints_.get(goodMatch.queryIdx).pt);
}
// System.out.println("------------------------------------------");
// System.out.println(goodNewKeypoints);
// System.out.println(goodOldKeypoints);
if (goodMatches.size() > 20) {
Mat goodNewPoints = Converters.vector_Point2f_to_Mat(goodNewKeypoints);
MatOfPoint2f originalNewPoints = new MatOfPoint2f();
Core.transform(goodNewPoints, originalNewPoints, Imgproc.getRotationMatrix2D(new Point(size.width / 2, size.height / 2), -angle, 1));
List<Point> shiftPoints = new ArrayList<>();
Converters.Mat_to_vector_Point2f(originalNewPoints, shiftPoints);
for (int i = 0; i < shiftPoints.size(); i++) {
double x = shiftPoints.get(i).x + crop * size.width / (1 - 2 * crop);
double y = shiftPoints.get(i).y + crop * size.height / (1 - 2 * crop);
shiftPoints.set(i, new Point(x, y));
}
Mat homography = Calib3d.findHomography(new MatOfPoint2f(shiftPoints.stream().toArray(Point[]::new)), new MatOfPoint2f(goodOldKeypoints.stream().toArray(Point[]::new)), Calib3d.RANSAC, 10);
Mat mask = new Mat(new Size(size.width / (1 - 2 * crop), size.height / (1 - 2 * crop)), CvType.CV_8UC1, new Scalar(255));
Mat maskWarpped = new Mat();
Imgproc.warpPerspective(mask, maskWarpped, homography, size);
Mat tmp = new Mat();
Imgproc.warpPerspective(frame, tmp, homography, size, Imgproc.INTER_LINEAR, Core.BORDER_REPLICATE, Scalar.all(255));
tmp.copyTo(stabilized, maskWarpped);
return new Img(stabilized, false);
}
return null;
}
private MatOfKeyPoint detect(Mat src) {
try (Img img = new Img(src);
Img adaptativThreshold = img.cvtColor(Imgproc.COLOR_BGR2GRAY).adaptativeThresHold(255, Imgproc.ADAPTIVE_THRESH_MEAN_C, Imgproc.THRESH_BINARY_INV, 17, 9);
Img closed = adaptativThreshold.morphologyEx(Imgproc.MORPH_CLOSE, Imgproc.MORPH_ELLIPSE, new Size(5, 5));) {
List<KeyPoint> keyPoints = new ArrayList<>();
List<MatOfPoint> contours = new ArrayList<>();
Imgproc.findContours(closed.getSrc(), contours, new Mat(), Imgproc.RETR_LIST, Imgproc.CHAIN_APPROX_SIMPLE);
double minArea = 100;
List<Rect> rects = contours.stream().filter(contour -> Imgproc.contourArea(contour) > minArea).map(contour -> Imgproc.boundingRect(contour)).collect(Collectors.toList());
rects.forEach(rect -> {
keyPoints.add(new KeyPoint(Double.valueOf(rect.tl().x).floatValue(), Double.valueOf(rect.tl().y).floatValue(), 6));
keyPoints.add(new KeyPoint(Double.valueOf(rect.tl().x).floatValue(), Double.valueOf(rect.br().y).floatValue(), 6));
keyPoints.add(new KeyPoint(Double.valueOf(rect.br().x).floatValue(), Double.valueOf(rect.tl().y).floatValue(), 6));
keyPoints.add(new KeyPoint(Double.valueOf(rect.br().x).floatValue(), Double.valueOf(rect.br().y).floatValue(), 6));
});
return new MatOfKeyPoint(keyPoints.stream().toArray(KeyPoint[]::new));
}
}
private Img deskew(Mat frame, double[] angle) {
try (Img img = new Img(frame);
Img adaptativThreshold = img.cvtColor(Imgproc.COLOR_BGR2GRAY).adaptativeThresHold(255, Imgproc.ADAPTIVE_THRESH_MEAN_C, Imgproc.THRESH_BINARY_INV, 17, 9);
Img closed = adaptativThreshold.morphologyEx(Imgproc.MORPH_CLOSE, Imgproc.MORPH_ELLIPSE, new Size(5, 5));) {
angle[0] = detection_contours(frame, closed.getSrc());
Mat matrix = Imgproc.getRotationMatrix2D(new Point(frame.width() / 2, frame.height() / 2), angle[0], 1);
Mat rotated = new Mat();
Imgproc.warpAffine(frame, rotated, matrix, new Size(frame.size().width, frame.size().height));
Img result = new Img(new Mat(rotated,
new Rect(Double.valueOf(rotated.width() * crop).intValue(), Double.valueOf(rotated.height() * crop).intValue(), Double.valueOf(rotated.width() * (1 - 2 * crop)).intValue(), Double.valueOf(rotated.height() * (1 - 2 * crop)).intValue())),
true);
matrix.release();
rotated.release();
return result;
}
}
private double detection_contours(Mat frame, Mat dilated) {
List<MatOfPoint> contours = new ArrayList<>();
Imgproc.findContours(dilated, contours, new Mat(), Imgproc.RETR_LIST, Imgproc.CHAIN_APPROX_SIMPLE);
double minArea = 100;
Predicate<RotatedRect> filter = rect -> rect.center.x > Double.valueOf(frame.width() * crop).intValue() && rect.center.y > Double.valueOf(frame.height() * crop).intValue() && rect.center.x < Double.valueOf(frame.width() * (1 - crop)).intValue()
&& rect.center.y < Double.valueOf(frame.height() * (1 - crop)).intValue();
List<RotatedRect> rotatedRects = contours.stream().filter(contour -> Imgproc.contourArea(contour) > minArea).map(contour -> Imgproc.minAreaRect(new MatOfPoint2f(contour.toArray()))).filter(filter).collect(Collectors.toList());
double mean = 0;
for (RotatedRect rotatedRect : rotatedRects) {
if (rotatedRect.angle < -45.) {
rotatedRect.angle += 90.0;
double tmp = rotatedRect.size.width;
rotatedRect.size.width = rotatedRect.size.height;
rotatedRect.size.height = tmp;
}
mean += rotatedRect.angle;
}
final double average = mean / rotatedRects.size();
List<RotatedRect> goodRects = rotatedRects.stream().filter(rotatedRect -> Math.abs(rotatedRect.angle - average) < 5).collect(Collectors.toList());
double goodRectsMean = 0;
for (RotatedRect rotatedRect : goodRects) {
goodRectsMean += rotatedRect.angle;
}
final double goodAverage = goodRectsMean / goodRects.size();
goodRects.forEach(rotatedRect -> rotatedRect.angle = goodAverage);
// System.out.println(average);
// System.out.println(goodAverage);
goodRects.forEach(rotatedRect -> {
Point[] result = new Point[4];
rotatedRect.points(result);
List<MatOfPoint> mof = Collections.singletonList(new MatOfPoint(new MatOfPoint(result)));
// Imgproc.drawContours(frame, mof, 0, new Scalar(0, 255, 0), 1);
// Imgproc.drawContours(dilated, mof, 0, new Scalar(255), 1);
});
return goodAverage;
}
// public void detection_deskew_contours(Mat frame, Mat dilated) {
// List<MatOfPoint> contours = new ArrayList<>();
// Imgproc.findContours(dilated, contours, new Mat(), Imgproc.RETR_LIST, Imgproc.CHAIN_APPROX_SIMPLE);
// double minArea = 100;
// List<Rect> rects = contours.stream().filter(contour -> Imgproc.contourArea(contour) > minArea).map(contour -> Imgproc.boundingRect(contour)).collect(Collectors.toList());
// rects.forEach(rect -> Imgproc.rectangle(frame, rect.tl(), rect.br(), new Scalar(0, 255, 0), 1));
// }
@Override
public void stop() throws Exception {
timer.shutdown();
capture.release();
super.stop();
}
}