public void performMatch() {
// create feature detectors and feature extractors
FeatureDetector orbDetector = FeatureDetector.create(FeatureDetector.ORB);
DescriptorExtractor orbExtractor = DescriptorExtractor.create(DescriptorExtractor.ORB);
// set the keypoints
keyPointImg = new MatOfKeyPoint();
orbDetector.detect(imgGray, keyPointImg);
MatOfKeyPoint keyPointTempl = new MatOfKeyPoint();
orbDetector.detect(templGray, keyPointTempl);
// get the descriptions
descImg = new Mat(image.size(), image.type());
orbExtractor.compute(imgGray, keyPointImg, descImg);
Mat descTempl = new Mat(template.size(), template.type());
orbExtractor.compute(templGray, keyPointTempl, descTempl);
// perform matching
matches = new MatOfDMatch();
DescriptorMatcher matcher = DescriptorMatcher.create(DescriptorMatcher.BRUTEFORCE_HAMMING);
matcher.match(descImg, descTempl, matches);
Log.i("perform match result", matches.size().toString());
}
public void testClear() {
matcher.add(Arrays.asList(new Mat()));
matcher.clear();
assertTrue(matcher.empty());
}
public void testCloneBoolean() {
matcher.add(Arrays.asList(new Mat()));
DescriptorMatcher cloned = matcher.clone(true);
assertNotNull(cloned);
assertTrue(cloned.empty());
}
public void testMatchMatListOfDMatch() {
Mat train = getTrainDescriptors();
Mat query = getQueryDescriptors();
MatOfDMatch matches = new MatOfDMatch();
matcher.add(Arrays.asList(train));
matcher.match(query, matches);
assertArrayDMatchEquals(truth, matches.toArray(), EPS);
}
public void testGetTrainDescriptors() {
Mat train = new Mat(1, 1, CvType.CV_8U, new Scalar(123));
Mat truth = train.clone();
matcher.add(Arrays.asList(train));
List<Mat> descriptors = matcher.getTrainDescriptors();
assertEquals(1, descriptors.size());
assertMatEqual(truth, descriptors.get(0));
}
public void testMatchMatListOfDMatchListOfMat() {
Mat train = getTrainDescriptors();
Mat query = getQueryDescriptors();
Mat mask = getMaskImg();
MatOfDMatch matches = new MatOfDMatch();
matcher.add(Arrays.asList(train));
matcher.match(query, matches, Arrays.asList(mask));
assertListDMatchEquals(Arrays.asList(truth[0], truth[1]), matches.toList(), EPS);
}
public void testRead() {
String filename = OpenCVTestRunner.getTempFileName("yml");
writeFile(filename, "%YAML:1.0\n");
matcher.read(filename);
assertTrue(true); // BruteforceMatcher has no settings
}
public void testWrite() {
String filename = OpenCVTestRunner.getTempFileName("yml");
matcher.write(filename);
String truth = "%YAML:1.0\n";
assertEquals(truth, readFile(filename));
}
public void onCameraViewStarted(int width, int height) {
try {
detectThread = null;
if (detectThread == null) {
detectThread = new DetectThread();
thread = new Thread(detectThread);
thread.setPriority(Thread.MIN_PRIORITY);
thread.start();
}
// -- Step 1: Detect the keypoints using SURF Detector
featureDetector = FeatureDetector.create(FeatureDetector.FAST);
// -- Step 2: Calculate descriptors (feature vectors)
extractor = DescriptorExtractor.create(DescriptorExtractor.FREAK);
// -- Step 3: Matching descriptor vectors using FLANN matcher
matcher = DescriptorMatcher.create(DescriptorMatcher.BRUTEFORCE_L1);
img_object = Utils.loadResource(this, R.drawable.cardobj, Highgui.CV_LOAD_IMAGE_GRAYSCALE);
img_scene = new Mat();
// img_scene = Utils.loadResource(this, R.drawable.cardscene,
// Highgui.CV_LOAD_IMAGE_GRAYSCALE);
matches = new MatOfDMatch();
listGoodMatches = new ArrayList<>();
keypoints_object = new MatOfKeyPoint();
keypoints_scene = new MatOfKeyPoint();
descriptors_object = new Mat();
descriptors_scene = new Mat();
obj = new MatOfPoint2f();
scene = new MatOfPoint2f();
H = new Mat();
obj_corners = new MatOfPoint2f();
scene_corners = new MatOfPoint2f();
min_dist = 9999999;
p0 = new Point(0, 0);
p1 = new Point(0, 0);
p2 = new Point(0, 0);
p3 = new Point(0, 0);
good_matches = new MatOfDMatch();
listPointScene = new ArrayList<>();
} catch (Exception ex) {
}
}
protected void setUp() throws Exception {
super.setUp();
matcher = DescriptorMatcher.create(DescriptorMatcher.BRUTEFORCE_HAMMINGLUT);
matSize = 100;
truth =
new DMatch[] {
new DMatch(0, 0, 0, 51),
new DMatch(1, 2, 0, 42),
new DMatch(2, 1, 0, 40),
new DMatch(3, 3, 0, 53)
};
}
public void testMatchMatMatListOfDMatch() {
Mat train = getTrainDescriptors();
Mat query = getQueryDescriptors();
MatOfDMatch matches = new MatOfDMatch();
matcher.match(query, train, matches);
/*
OpenCVTestRunner.Log("matches found: " + matches.size());
for (DMatch m : matches.toArray())
OpenCVTestRunner.Log(m.toString());
*/
assertArrayDMatchEquals(truth, matches.toArray(), EPS);
}
public void map(Text key, Text value, Context context)
throws InterruptedException, IOException {
String filename = key.toString();
String json = value.toString();
// Make sure the input is valid
if (!(filename.isEmpty() || json.isEmpty())) {
// Change the json-type feature to Mat-type feature
Mat descriptor = json2mat(json);
if (descriptor != null) {
// Read the query feature from the cache in Hadoop
Mat query_features;
String pathStr = context.getConfiguration().get("featureFilePath");
FileSystem fs = FileSystem.get(context.getConfiguration());
FSDataInputStream fsDataInputStream = fs.open(new Path(pathStr));
StringBuilder sb = new StringBuilder();
// Use a buffer to read the query_feature
int remain = fsDataInputStream.available();
while (remain > 0) {
int read;
byte[] buf = new byte[BUF_SIZE];
read = fsDataInputStream.read(buf, fsDataInputStream.available() - remain, BUF_SIZE);
sb.append(new String(buf, 0, read, StandardCharsets.UTF_8));
remain = remain - read;
System.out.println("remain:" + remain + "\tread:" + read + "\tsb.size:" + sb.length());
}
// Read the query_feature line by line
// Scanner sc = new Scanner(fsDataInputStream, "UTF-8");
// StringBuilder sb = new StringBuilder();
// while (sc.hasNextLine()) {
// sb.append(sc.nextLine());
// }
// String query_json = sb.toString();
// String query_json = new String(buf, StandardCharsets.UTF_8);
String query_json = sb.toString();
fsDataInputStream.close();
query_features = json2mat(query_json);
// Get the similarity of the current database image against the query image
DescriptorMatcher matcher = DescriptorMatcher.create(DescriptorMatcher.FLANNBASED);
MatOfDMatch matches = new MatOfDMatch();
// Ensure the two features have same length of cols (the feature extracted are all 128
// cols(at least in this case))
if (query_features.cols() == descriptor.cols()) {
matcher.match(query_features, descriptor, matches);
DMatch[] dMatches = matches.toArray();
// Calculate the max/min distances
// double max_dist = Double.MAX_VALUE;
// double min_dist = Double.MIN_VALUE;
double max_dist = 0;
double min_dist = 100;
for (int i = 0; i < dMatches.length; i++) {
double dist = dMatches[i].distance;
if (min_dist > dist) min_dist = dist;
if (max_dist < dist) max_dist = dist;
}
// Only distances ≤ threshold are good matches
double threshold = max_dist * THRESHOLD_FACTOR;
// double threshold = min_dist * 2;
LinkedList<DMatch> goodMatches = new LinkedList<DMatch>();
for (int i = 0; i < dMatches.length; i++) {
if (dMatches[i].distance <= threshold) {
goodMatches.addLast(dMatches[i]);
}
}
// Get the ratio of good_matches to all_matches
double ratio = (double) goodMatches.size() / (double) dMatches.length;
System.out.println("*** current_record_filename:" + filename + " ***");
System.out.println("feature:" + descriptor + "\nquery_feature:" + query_features);
System.out.println(
"min_dist of keypoints:" + min_dist + " max_dist of keypoints:" + max_dist);
System.out.println(
"total_matches:" + dMatches.length + "\tgood_matches:" + goodMatches.size());
// System.out.println("type:" + descriptor.type() + " channels:" +
// descriptor.channels() + " rows:" + descriptor.rows() + " cols:" + descriptor.cols());
// System.out.println("qtype:" + query_features.type() + "
// qchannels:" + query_features.channels() + " qrows:" + query_features.rows() + "
// qcols:" + query_features.cols());
System.out.println();
if (ratio > PERCENTAGE_THRESHOLD) {
// Key:1 Value:filename|ratio
context.write(ONE, new Text(filename + "|" + ratio));
// context.write(ONE, new Text(filename + "|" +
// String.valueOf(goodMatches.size())));
}
} else {
System.out.println("The size of the features are not equal");
}
} else {
// a null pointer, do nothing
System.out.println("A broken/null feature:" + filename);
System.out.println();
}
}
}
public void testIsMaskSupported() {
assertTrue(matcher.isMaskSupported());
}
public Template performMatches(Map<String, Template> templates) {
// create feature detectors and feature extractors
FeatureDetector orbDetector = FeatureDetector.create(FeatureDetector.ORB);
DescriptorExtractor orbExtractor = DescriptorExtractor.create(DescriptorExtractor.ORB);
MatOfKeyPoint keyPointImgT;
Mat descImgT;
// set the keypoints
keyPointImgT = new MatOfKeyPoint();
orbDetector.detect(imgGray, keyPointImgT);
descImgT = new Mat(image.size(), image.type());
orbExtractor.compute(imgGray, keyPointImgT, descImgT);
Template best = null;
matches = null;
Map.Entry<String, Template> maxEntry = null;
// MatOfDMatch matches = new MatOfDMatch();
for (Map.Entry<String, Template> entry : templates.entrySet()) {
MatOfKeyPoint keyPointTempl = null;
Mat descTempl = null;
Mat tGray = null;
Template t = entry.getValue();
if (null == t.getTemplGray() || null == t.getDescTempl() || null == t.getKeyPointTempl()) {
// read image from stored data
Mat templ = readImgFromFile(t.getTemplName());
tGray = new Mat(templ.size(), templ.type());
Imgproc.cvtColor(templ, tGray, Imgproc.COLOR_BGRA2GRAY);
keyPointTempl = new MatOfKeyPoint();
orbDetector.detect(tGray, keyPointTempl);
descTempl = new Mat(templ.size(), templ.type());
orbExtractor.compute(tGray, keyPointTempl, descTempl);
t.setKeyPointTempl(keyPointTempl);
t.setDescTempl(descTempl);
} else {
descTempl = t.getDescTempl();
}
MatOfDMatch matchWithT = new MatOfDMatch();
DescriptorMatcher matcher = DescriptorMatcher.create(DescriptorMatcher.BRUTEFORCE_HAMMING);
// matcher.radiusMatch(descImgT, descTempl, matchWithT,200);//
matcher.match(descImgT, descTempl, matchWithT);
List<DMatch> matchList = matchWithT.toList();
// float min = Float.MAX_VALUE;
// float max = Float.MIN_VALUE;
// for(int i=0;i<matchList.size();i++){
// min = matchList.get(i).distance<min?matchList.get(i).distance:min;
// max = matchList.get(i).distance>max?matchList.get(i).distance:max;
// }
// Log.i("min distance","min distance is::"+min+"max
// distance::"+max+"size::"+matchList.size());
// Collections.sort(matchList, new Comparator<DMatch>() {
// @Override
// public int compare(DMatch o1, DMatch o2) {
// if (o1.distance < o2.distance)
// return -1;
// if (o1.distance > o2.distance)
// return 1;
// return 0;
// }
// });
float ratio = -1;
if (matchList.size() > 0) ratio = findMinTwoRatio(matchList);
if (ratio > 0.8 || ratio == -1) continue;
Log.i("match", "ratio::" + ratio);
// Todo:revisit logic
if (matches == null || (matchWithT.size().height > matches.size().height)) {
matches = matchWithT;
keyPointImg = keyPointImgT;
descImg = descImgT;
best = t;
}
}
// Log.i("perform match result", matches.size().toString());
return best;
}
private void detectObject() {
readLock.lock();
featureDetector.detect(img_scene, keypoints_scene);
featureDetector.detect(img_object, keypoints_object);
extractor.compute(img_object, keypoints_object, descriptors_object);
extractor.compute(img_scene, keypoints_scene, descriptors_scene);
readLock.unlock();
if (!descriptors_scene.empty()) {
matcher.match(descriptors_object, descriptors_scene, matches);
// readLock.unlock();
//
listMatches = matches.toList();
int size = descriptors_object.rows();
// -- Quick calculation of max and min distances between keypoints
for (int i = 0; i < size; i++) {
double dist = listMatches.get(i).distance;
if (dist < min_dist) {
min_dist = dist;
}
}
Log.e("Min", min_dist + "");
threeMinDist = 3 * min_dist;
listGoodMatches.removeAll(listGoodMatches);
for (int i = 0; i < size; i++) {
DMatch dMatch = listMatches.get(i);
float distance = dMatch.distance;
if (distance < threeMinDist) {
listGoodMatches.add(dMatch);
}
}
// good_matches.fromList(listGoodMatches);
Log.e("Matches", listMatches.size() + "");
Log.e("Good Matches", listGoodMatches.size() + "");
//
if (listGoodMatches.size() > 4) {
Point pointObj[] = new Point[listGoodMatches.size()];
Point pointScene[] = new Point[listGoodMatches.size()];
listKeyPointObject = keypoints_object.toList();
listKeyPointScene = keypoints_scene.toList();
// listPointScene.removeAll(listPointScene);
for (int i = 0; i < listGoodMatches.size(); i++) {
// -- Get the keypoints from the good matches
pointObj[i] = listKeyPointObject.get(listGoodMatches.get(i).queryIdx).pt;
pointScene[i] = listKeyPointScene.get(listGoodMatches.get(i).trainIdx).pt;
// listPointScene.add(listKeyPointScene.get(listGoodMatches.get(i).trainIdx).pt);
}
obj.fromArray(pointObj);
scene.fromArray(pointScene);
Log.e("Before findHomography", "");
H = Calib3d.findHomography(obj, scene, Calib3d.RANSAC, 9);
Log.e("AFTERRR findHomography", "");
pointObjConners[0] = new Point(0, 0);
pointObjConners[1] = new Point(img_object.cols(), 0);
pointObjConners[2] = new Point(img_object.cols(), img_object.rows());
pointObjConners[3] = new Point(0, img_object.rows());
obj_corners.fromArray(pointObjConners);
Core.perspectiveTransform(obj_corners, scene_corners, H);
p0 = new Point(scene_corners.toList().get(0).x, scene_corners.toList().get(0).y + 0);
p1 = new Point(scene_corners.toList().get(1).x, scene_corners.toList().get(1).y + 0);
p2 = new Point(scene_corners.toList().get(2).x, scene_corners.toList().get(2).y + 0);
p3 = new Point(scene_corners.toList().get(3).x, scene_corners.toList().get(3).y + 0);
Log.e("POINT THREAD", p0.toString() + p1.toString() + p2.toString() + p3.toString());
Log.e("detect ok", "detect ok");
}
} else {
Log.e("No descritor", "No descritor");
// readLock.unlock();
}
}
public void testAdd() {
matcher.add(Arrays.asList(new Mat()));
assertFalse(matcher.empty());
}
public void testEmpty() {
assertTrue(matcher.empty());
}
public void testTrain() {
matcher.train(); // BruteforceMatcher does not need to train
}
