Jak sprawdzić, czy dwa obrazy są podobne lub nie korzystają z openCV w java?

Question

muszę sprawdzić, czy dwa obrazy są podobne lub nie w java za pomocą OpenCV, używam OpenCV za to i za pomocą ORB

Oto moje główne klasy

System.out.println("Welcome to OpenCV " + Core.VERSION); 
System.loadLibrary(Core.NATIVE_LIBRARY_NAME);()); 
System.out.println(System.getProperty("user.dir")); 
File f1 = new File(System.getProperty("user.dir") + "\\test.jpg"); 
File f2 = new File(System.getProperty("user.dir") + "\\test2.jpg"); 
MatchingDemo2 m = new MatchingDemo2(); 
m.mth(f1.getAbsolutePath(), f2.getAbsolutePath()); 

I tu jest moje MatchingDemo2. java plik

public class MatchingDemo2 { 

public void mth(String inFile, String templateFile){ 
     FeatureDetector detector = FeatureDetector.create(FeatureDetector.ORB); 
     //Create descriptors 

     //first image 
     // generate descriptors 

     //second image 
     // generate descriptors 



     System.out.println("size " + matches.size()); 
     //HOW DO I KNOW IF IMAGES MATCHED OR NOT ???? 

     //THIS CODE IS FOR CONNECTIONS BUT I AM NOT ABLE TO DO IT 
     //feature and connection colors 
     Scalar RED = new Scalar(255,0,0); 
     Scalar GREEN = new Scalar(0,255,0); 
     //output image 
     Mat outputImg = new Mat(); 
     MatOfByte drawnMatches = new MatOfByte(); 
     //this will draw all matches, works fine 
     Features2d.drawMatches(img1, keypoints1, img2, keypoints2, matches, 
       outputImg, GREEN, RED, drawnMatches, Features2d.NOT_DRAW_SINGLE_POINTS); 
     int DIST_LIMIT = 80; 
     List<DMatch> matchList = matches.toList(); 
     List<DMatch> matches_final = new ArrayList<DMatch>(); 
     for(int i=0; i<matchList.size(); i++){ 
      if(matchList.get(i).distance <= DIST_LIMIT){ 
       matches_final.add(matches.toList().get(i)); 
      } 
     } 

     MatOfDMatch matches_final_mat = new MatOfDMatch(); 
     matches_final_mat.fromList(matches_final); 
     for(int i=0; i< matches_final.size(); i++){ 
      System.out.println("Good Matchs "+ matches_final.get(i)); 
     } 
} 
} 

Ale kiedy sprawdzić Dobre Matchs ja dostać ten

size 1x328 
Good Matchs DMatch [queryIdx=0, trainIdx=93, imgIdx=0, distance=52.0] 
Good Matchs DMatch [queryIdx=1, trainIdx=173, imgIdx=0, distance=57.0] 
Good Matchs DMatch [queryIdx=2, trainIdx=92, imgIdx=0, distance=53.0] 
Good Matchs DMatch [queryIdx=3, trainIdx=80, imgIdx=0, distance=26.0] 
Good Matchs DMatch [queryIdx=5, trainIdx=164, imgIdx=0, distance=40.0] 
Good Matchs DMatch [queryIdx=6, trainIdx=228, imgIdx=0, distance=53.0] 
Good Matchs DMatch [queryIdx=7, trainIdx=179, imgIdx=0, distance=14.0] 
Good Matchs DMatch [queryIdx=8, trainIdx=78, imgIdx=0, distance=61.0] 
Good Matchs DMatch [queryIdx=9, trainIdx=166, imgIdx=0, distance=69.0] 
Good Matchs DMatch [queryIdx=10, trainIdx=74, imgIdx=0, distance=65.0] 
Good Matchs DMatch [queryIdx=11, trainIdx=245, imgIdx=0, distance=38.0] 
Good Matchs DMatch [queryIdx=12, trainIdx=120, imgIdx=0, distance=66.0] 
Good Matchs DMatch [queryIdx=13, trainIdx=244, imgIdx=0, distance=41.0] 
Good Matchs DMatch [queryIdx=14, trainIdx=67, imgIdx=0, distance=50.0] 
Good Matchs DMatch [queryIdx=15, trainIdx=185, imgIdx=0, distance=55.0] 
Good Matchs DMatch [queryIdx=16, trainIdx=97, imgIdx=0, distance=21.0] 
Good Matchs DMatch [queryIdx=17, trainIdx=172, imgIdx=0, distance=67.0] 
Good Matchs DMatch [queryIdx=18, trainIdx=354, imgIdx=0, distance=67.0] 
Good Matchs DMatch [queryIdx=19, trainIdx=302, imgIdx=0, distance=53.0] 
Good Matchs DMatch [queryIdx=20, trainIdx=176, imgIdx=0, distance=59.0] 
Good Matchs DMatch [queryIdx=21, trainIdx=60, imgIdx=0, distance=66.0] 
Good Matchs DMatch [queryIdx=22, trainIdx=72, imgIdx=0, distance=62.0] 
Good Matchs DMatch [queryIdx=23, trainIdx=63, imgIdx=0, distance=54.0] 
Good Matchs DMatch [queryIdx=24, trainIdx=176, imgIdx=0, distance=63.0] 
Good Matchs DMatch [queryIdx=25, trainIdx=49, imgIdx=0, distance=58.0] 
Good Matchs DMatch [queryIdx=26, trainIdx=77, imgIdx=0, distance=52.0] 
Good Matchs DMatch [queryIdx=27, trainIdx=302, imgIdx=0, distance=59.0] 
Good Matchs DMatch [queryIdx=28, trainIdx=265, imgIdx=0, distance=71.0] 
Good Matchs DMatch [queryIdx=29, trainIdx=67, imgIdx=0, distance=49.0] 
Good Matchs DMatch [queryIdx=30, trainIdx=302, imgIdx=0, distance=64.0] 
Good Matchs DMatch [queryIdx=31, trainIdx=265, imgIdx=0, distance=65.0] 
Good Matchs DMatch [queryIdx=32, trainIdx=73, imgIdx=0, distance=73.0] 
Good Matchs DMatch [queryIdx=33, trainIdx=67, imgIdx=0, distance=55.0] 
Good Matchs DMatch [queryIdx=34, trainIdx=283, imgIdx=0, distance=63.0] 
Good Matchs DMatch [queryIdx=35, trainIdx=145, imgIdx=0, distance=60.0] 
Good Matchs DMatch [queryIdx=36, trainIdx=71, imgIdx=0, distance=54.0] 
Good Matchs DMatch [queryIdx=37, trainIdx=167, imgIdx=0, distance=67.0] 
Good Matchs DMatch [queryIdx=38, trainIdx=94, imgIdx=0, distance=60.0] 
Good Matchs DMatch [queryIdx=39, trainIdx=88, imgIdx=0, distance=68.0] 
Good Matchs DMatch [queryIdx=40, trainIdx=88, imgIdx=0, distance=69.0] 
Good Matchs DMatch [queryIdx=41, trainIdx=179, imgIdx=0, distance=28.0] 
Good Matchs DMatch [queryIdx=42, trainIdx=64, imgIdx=0, distance=61.0] 
Good Matchs DMatch [queryIdx=43, trainIdx=223, imgIdx=0, distance=71.0] 
Good Matchs DMatch [queryIdx=44, trainIdx=80, imgIdx=0, distance=30.0] 
Good Matchs DMatch [queryIdx=45, trainIdx=196, imgIdx=0, distance=59.0] 
Good Matchs DMatch [queryIdx=46, trainIdx=52, imgIdx=0, distance=65.0] 
Good Matchs DMatch [queryIdx=47, trainIdx=93, imgIdx=0, distance=60.0] 
Good Matchs DMatch [queryIdx=48, trainIdx=187, imgIdx=0, distance=49.0] 
Good Matchs DMatch [queryIdx=49, trainIdx=179, imgIdx=0, distance=50.0] 
Good Matchs DMatch [queryIdx=50, trainIdx=283, imgIdx=0, distance=59.0] 
Good Matchs DMatch [queryIdx=51, trainIdx=171, imgIdx=0, distance=61.0] 
Good Matchs DMatch [queryIdx=52, trainIdx=302, imgIdx=0, distance=60.0] 
Good Matchs DMatch [queryIdx=53, trainIdx=67, imgIdx=0, distance=59.0] 
Good Matchs DMatch [queryIdx=54, trainIdx=15, imgIdx=0, distance=59.0] 
Good Matchs DMatch [queryIdx=55, trainIdx=173, imgIdx=0, distance=66.0] 
Good Matchs DMatch [queryIdx=56, trainIdx=302, imgIdx=0, distance=56.0] 
Good Matchs DMatch [queryIdx=57, trainIdx=47, imgIdx=0, distance=56.0] 
Good Matchs DMatch [queryIdx=58, trainIdx=187, imgIdx=0, distance=58.0] 
Good Matchs DMatch [queryIdx=59, trainIdx=344, imgIdx=0, distance=64.0] 
Good Matchs DMatch [queryIdx=60, trainIdx=164, imgIdx=0, distance=65.0] 
Good Matchs DMatch [queryIdx=61, trainIdx=125, imgIdx=0, distance=50.0] 
Good Matchs DMatch [queryIdx=62, trainIdx=77, imgIdx=0, distance=72.0] 
Good Matchs DMatch [queryIdx=63, trainIdx=22, imgIdx=0, distance=79.0] 
Good Matchs DMatch [queryIdx=64, trainIdx=82, imgIdx=0, distance=69.0] 
Good Matchs DMatch [queryIdx=65, trainIdx=93, imgIdx=0, distance=63.0] 
Good Matchs DMatch [queryIdx=66, trainIdx=241, imgIdx=0, distance=35.0] 
Good Matchs DMatch [queryIdx=67, trainIdx=80, imgIdx=0, distance=18.0] 
Good Matchs DMatch [queryIdx=68, trainIdx=179, imgIdx=0, distance=20.0] 
Good Matchs DMatch [queryIdx=69, trainIdx=242, imgIdx=0, distance=50.0] 
Good Matchs DMatch [queryIdx=70, trainIdx=80, imgIdx=0, distance=22.0] 
Good Matchs DMatch [queryIdx=71, trainIdx=179, imgIdx=0, distance=19.0] 
Good Matchs DMatch [queryIdx=72, trainIdx=92, imgIdx=0, distance=59.0] 
Good Matchs DMatch [queryIdx=73, trainIdx=94, imgIdx=0, distance=56.0] 
Good Matchs DMatch [queryIdx=74, trainIdx=173, imgIdx=0, distance=49.0] 
Good Matchs DMatch [queryIdx=75, trainIdx=94, imgIdx=0, distance=59.0] 
Good Matchs DMatch [queryIdx=76, trainIdx=94, imgIdx=0, distance=48.0] 
Good Matchs DMatch [queryIdx=77, trainIdx=92, imgIdx=0, distance=52.0] 
Good Matchs DMatch [queryIdx=78, trainIdx=80, imgIdx=0, distance=20.0] 
Good Matchs DMatch [queryIdx=80, trainIdx=119, imgIdx=0, distance=61.0] 
Good Matchs DMatch [queryIdx=81, trainIdx=228, imgIdx=0, distance=47.0] 
Good Matchs DMatch [queryIdx=82, trainIdx=179, imgIdx=0, distance=14.0] 
Good Matchs DMatch [queryIdx=83, trainIdx=227, imgIdx=0, distance=52.0] 
Good Matchs DMatch [queryIdx=84, trainIdx=84, imgIdx=0, distance=57.0] 
Good Matchs DMatch [queryIdx=85, trainIdx=245, imgIdx=0, distance=40.0] 
Good Matchs DMatch [queryIdx=86, trainIdx=58, imgIdx=0, distance=64.0] 
Good Matchs DMatch [queryIdx=87, trainIdx=14, imgIdx=0, distance=62.0] 
Good Matchs DMatch [queryIdx=88, trainIdx=187, imgIdx=0, distance=61.0] 
Good Matchs DMatch [queryIdx=89, trainIdx=185, imgIdx=0, distance=57.0] 
Good Matchs DMatch [queryIdx=90, trainIdx=178, imgIdx=0, distance=25.0] 
Good Matchs DMatch [queryIdx=91, trainIdx=220, imgIdx=0, distance=56.0] 
Good Matchs DMatch [queryIdx=92, trainIdx=205, imgIdx=0, distance=59.0] 
Good Matchs DMatch [queryIdx=93, trainIdx=60, imgIdx=0, distance=61.0] 
Good Matchs DMatch [queryIdx=94, trainIdx=44, imgIdx=0, distance=59.0] 
Good Matchs DMatch [queryIdx=95, trainIdx=16, imgIdx=0, distance=72.0] 
Good Matchs DMatch [queryIdx=96, trainIdx=157, imgIdx=0, distance=63.0] 
Good Matchs DMatch [queryIdx=97, trainIdx=135, imgIdx=0, distance=52.0] 
Good Matchs DMatch [queryIdx=98, trainIdx=60, imgIdx=0, distance=61.0] 
Good Matchs DMatch [queryIdx=99, trainIdx=344, imgIdx=0, distance=60.0] 
Good Matchs DMatch [queryIdx=100, trainIdx=77, imgIdx=0, distance=67.0] 
Good Matchs DMatch [queryIdx=101, trainIdx=95, imgIdx=0, distance=76.0] 
Good Matchs DMatch [queryIdx=102, trainIdx=72, imgIdx=0, distance=45.0] 
Good Matchs DMatch [queryIdx=103, trainIdx=134, imgIdx=0, distance=70.0] 
Good Matchs DMatch [queryIdx=104, trainIdx=154, imgIdx=0, distance=54.0] 
Good Matchs DMatch [queryIdx=105, trainIdx=208, imgIdx=0, distance=77.0] 
Good Matchs DMatch [queryIdx=106, trainIdx=73, imgIdx=0, distance=79.0] 
Good Matchs DMatch [queryIdx=107, trainIdx=72, imgIdx=0, distance=61.0] 
Good Matchs DMatch [queryIdx=108, trainIdx=64, imgIdx=0, distance=73.0] 
Good Matchs DMatch [queryIdx=109, trainIdx=72, imgIdx=0, distance=63.0] 
Good Matchs DMatch [queryIdx=110, trainIdx=365, imgIdx=0, distance=66.0] 
Good Matchs DMatch [queryIdx=111, trainIdx=148, imgIdx=0, distance=64.0] 
Good Matchs DMatch [queryIdx=112, trainIdx=81, imgIdx=0, distance=42.0] 
Good Matchs DMatch [queryIdx=113, trainIdx=56, imgIdx=0, distance=62.0] 
Good Matchs DMatch [queryIdx=114, trainIdx=162, imgIdx=0, distance=48.0] 
Good Matchs DMatch [queryIdx=115, trainIdx=56, imgIdx=0, distance=56.0] 
Good Matchs DMatch [queryIdx=116, trainIdx=120, imgIdx=0, distance=58.0] 
Good Matchs DMatch [queryIdx=117, trainIdx=72, imgIdx=0, distance=59.0] 
Good Matchs DMatch [queryIdx=118, trainIdx=92, imgIdx=0, distance=63.0] 
Good Matchs DMatch [queryIdx=119, trainIdx=131, imgIdx=0, distance=59.0] 
Good Matchs DMatch [queryIdx=120, trainIdx=72, imgIdx=0, distance=46.0] 
Good Matchs DMatch [queryIdx=121, trainIdx=74, imgIdx=0, distance=78.0] 
Good Matchs DMatch [queryIdx=122, trainIdx=94, imgIdx=0, distance=56.0] 
Good Matchs DMatch [queryIdx=123, trainIdx=72, imgIdx=0, distance=52.0] 
Good Matchs DMatch [queryIdx=124, trainIdx=134, imgIdx=0, distance=71.0] 
Good Matchs DMatch [queryIdx=125, trainIdx=72, imgIdx=0, distance=46.0] 
Good Matchs DMatch [queryIdx=126, trainIdx=15, imgIdx=0, distance=73.0] 
Good Matchs DMatch [queryIdx=127, trainIdx=72, imgIdx=0, distance=50.0] 
Good Matchs DMatch [queryIdx=128, trainIdx=93, imgIdx=0, distance=56.0] 
Good Matchs DMatch [queryIdx=129, trainIdx=68, imgIdx=0, distance=46.0] 
Good Matchs DMatch [queryIdx=130, trainIdx=205, imgIdx=0, distance=65.0] 
Good Matchs DMatch [queryIdx=131, trainIdx=187, imgIdx=0, distance=59.0] 
Good Matchs DMatch [queryIdx=132, trainIdx=72, imgIdx=0, distance=47.0] 
Good Matchs DMatch [queryIdx=133, trainIdx=220, imgIdx=0, distance=57.0] 
Good Matchs DMatch [queryIdx=134, trainIdx=289, imgIdx=0, distance=60.0] 
Good Matchs DMatch [queryIdx=135, trainIdx=82, imgIdx=0, distance=65.0] 
Good Matchs DMatch [queryIdx=136, trainIdx=93, imgIdx=0, distance=53.0] 
Good Matchs DMatch [queryIdx=137, trainIdx=244, imgIdx=0, distance=18.0] 
Good Matchs DMatch [queryIdx=138, trainIdx=244, imgIdx=0, distance=25.0] 
Good Matchs DMatch [queryIdx=139, trainIdx=92, imgIdx=0, distance=66.0] 
Good Matchs DMatch [queryIdx=140, trainIdx=244, imgIdx=0, distance=20.0] 
Good Matchs DMatch [queryIdx=141, trainIdx=93, imgIdx=0, distance=45.0] 
Good Matchs DMatch [queryIdx=142, trainIdx=93, imgIdx=0, distance=51.0] 
Good Matchs DMatch [queryIdx=143, trainIdx=94, imgIdx=0, distance=51.0] 
Good Matchs DMatch [queryIdx=144, trainIdx=94, imgIdx=0, distance=40.0] 
Good Matchs DMatch [queryIdx=145, trainIdx=93, imgIdx=0, distance=47.0] 
Good Matchs DMatch [queryIdx=146, trainIdx=244, imgIdx=0, distance=28.0] 
Good Matchs DMatch [queryIdx=147, trainIdx=172, imgIdx=0, distance=77.0] 
Good Matchs DMatch [queryIdx=148, trainIdx=170, imgIdx=0, distance=72.0] 
Good Matchs DMatch [queryIdx=149, trainIdx=261, imgIdx=0, distance=72.0] 
Good Matchs DMatch [queryIdx=150, trainIdx=228, imgIdx=0, distance=55.0] 
Good Matchs DMatch [queryIdx=151, trainIdx=179, imgIdx=0, distance=19.0] 
Good Matchs DMatch [queryIdx=152, trainIdx=227, imgIdx=0, distance=60.0] 
Good Matchs DMatch [queryIdx=153, trainIdx=107, imgIdx=0, distance=56.0] 
Good Matchs DMatch [queryIdx=154, trainIdx=174, imgIdx=0, distance=41.0] 
Good Matchs DMatch [queryIdx=155, trainIdx=283, imgIdx=0, distance=76.0] 
Good Matchs DMatch [queryIdx=156, trainIdx=254, imgIdx=0, distance=59.0] 
Good Matchs DMatch [queryIdx=157, trainIdx=185, imgIdx=0, distance=51.0] 
Good Matchs DMatch [queryIdx=158, trainIdx=178, imgIdx=0, distance=30.0] 
Good Matchs DMatch [queryIdx=159, trainIdx=278, imgIdx=0, distance=53.0] 
Good Matchs DMatch [queryIdx=160, trainIdx=91, imgIdx=0, distance=62.0] 
Good Matchs DMatch [queryIdx=161, trainIdx=148, imgIdx=0, distance=59.0] 
Good Matchs DMatch [queryIdx=162, trainIdx=157, imgIdx=0, distance=63.0] 
Good Matchs DMatch [queryIdx=163, trainIdx=373, imgIdx=0, distance=73.0] 
Good Matchs DMatch [queryIdx=164, trainIdx=226, imgIdx=0, distance=48.0] 
Good Matchs DMatch [queryIdx=165, trainIdx=278, imgIdx=0, distance=64.0] 
Good Matchs DMatch [queryIdx=166, trainIdx=283, imgIdx=0, distance=62.0] 
Good Matchs DMatch [queryIdx=167, trainIdx=196, imgIdx=0, distance=64.0] 
Good Matchs DMatch [queryIdx=168, trainIdx=344, imgIdx=0, distance=59.0] 
Good Matchs DMatch [queryIdx=169, trainIdx=157, imgIdx=0, distance=53.0] 
Good Matchs DMatch [queryIdx=170, trainIdx=144, imgIdx=0, distance=79.0] 
Good Matchs DMatch [queryIdx=171, trainIdx=154, imgIdx=0, distance=61.0] 
Good Matchs DMatch [queryIdx=172, trainIdx=211, imgIdx=0, distance=75.0] 
Good Matchs DMatch [queryIdx=173, trainIdx=279, imgIdx=0, distance=64.0] 
Good Matchs DMatch [queryIdx=174, trainIdx=211, imgIdx=0, distance=79.0] 
Good Matchs DMatch [queryIdx=175, trainIdx=220, imgIdx=0, distance=68.0] 
Good Matchs DMatch [queryIdx=176, trainIdx=218, imgIdx=0, distance=45.0] 
Good Matchs DMatch [queryIdx=177, trainIdx=289, imgIdx=0, distance=75.0] 
Good Matchs DMatch [queryIdx=178, trainIdx=223, imgIdx=0, distance=63.0] 
Good Matchs DMatch [queryIdx=179, trainIdx=57, imgIdx=0, distance=65.0] 
Good Matchs DMatch [queryIdx=180, trainIdx=36, imgIdx=0, distance=73.0] 
Good Matchs DMatch [queryIdx=181, trainIdx=111, imgIdx=0, distance=72.0] 
Good Matchs DMatch [queryIdx=182, trainIdx=93, imgIdx=0, distance=65.0] 
Good Matchs DMatch [queryIdx=183, trainIdx=137, imgIdx=0, distance=63.0] 
Good Matchs DMatch [queryIdx=184, trainIdx=157, imgIdx=0, distance=60.0] 
Good Matchs DMatch [queryIdx=185, trainIdx=72, imgIdx=0, distance=61.0] 
Good Matchs DMatch [queryIdx=186, trainIdx=172, imgIdx=0, distance=47.0] 
Good Matchs DMatch [queryIdx=187, trainIdx=279, imgIdx=0, distance=69.0] 
Good Matchs DMatch [queryIdx=188, trainIdx=72, imgIdx=0, distance=55.0] 
Good Matchs DMatch [queryIdx=189, trainIdx=96, imgIdx=0, distance=61.0] 
Good Matchs DMatch [queryIdx=190, trainIdx=220, imgIdx=0, distance=68.0] 
Good Matchs DMatch [queryIdx=191, trainIdx=93, imgIdx=0, distance=48.0] 
Good Matchs DMatch [queryIdx=192, trainIdx=279, imgIdx=0, distance=54.0] 
Good Matchs DMatch [queryIdx=193, trainIdx=157, imgIdx=0, distance=54.0] 
Good Matchs DMatch [queryIdx=194, trainIdx=91, imgIdx=0, distance=65.0] 
Good Matchs DMatch [queryIdx=195, trainIdx=278, imgIdx=0, distance=66.0] 
Good Matchs DMatch [queryIdx=196, trainIdx=220, imgIdx=0, distance=57.0] 
Good Matchs DMatch [queryIdx=197, trainIdx=74, imgIdx=0, distance=63.0] 
Good Matchs DMatch [queryIdx=198, trainIdx=93, imgIdx=0, distance=34.0] 
Good Matchs DMatch [queryIdx=199, trainIdx=81, imgIdx=0, distance=53.0] 
Good Matchs DMatch [queryIdx=200, trainIdx=93, imgIdx=0, distance=45.0] 
Good Matchs DMatch [queryIdx=201, trainIdx=90, imgIdx=0, distance=52.0] 
Good Matchs DMatch [queryIdx=202, trainIdx=93, imgIdx=0, distance=52.0] 
Good Matchs DMatch [queryIdx=203, trainIdx=94, imgIdx=0, distance=42.0] 
Good Matchs DMatch [queryIdx=204, trainIdx=93, imgIdx=0, distance=35.0] 
Good Matchs DMatch [queryIdx=205, trainIdx=94, imgIdx=0, distance=44.0] 
Good Matchs DMatch [queryIdx=206, trainIdx=90, imgIdx=0, distance=72.0] 
Good Matchs DMatch [queryIdx=207, trainIdx=179, imgIdx=0, distance=54.0] 
Good Matchs DMatch [queryIdx=208, trainIdx=92, imgIdx=0, distance=48.0] 
Good Matchs DMatch [queryIdx=209, trainIdx=91, imgIdx=0, distance=61.0] 
Good Matchs DMatch [queryIdx=210, trainIdx=119, imgIdx=0, distance=77.0] 
Good Matchs DMatch [queryIdx=211, trainIdx=227, imgIdx=0, distance=66.0] 
Good Matchs DMatch [queryIdx=212, trainIdx=186, imgIdx=0, distance=56.0] 
Good Matchs DMatch [queryIdx=213, trainIdx=96, imgIdx=0, distance=69.0] 
Good Matchs DMatch [queryIdx=214, trainIdx=77, imgIdx=0, distance=52.0] 
Good Matchs DMatch [queryIdx=215, trainIdx=372, imgIdx=0, distance=67.0] 
Good Matchs DMatch [queryIdx=216, trainIdx=334, imgIdx=0, distance=69.0] 
Good Matchs DMatch [queryIdx=217, trainIdx=278, imgIdx=0, distance=67.0] 
Good Matchs DMatch [queryIdx=218, trainIdx=325, imgIdx=0, distance=61.0] 
Good Matchs DMatch [queryIdx=219, trainIdx=188, imgIdx=0, distance=60.0] 
Good Matchs DMatch [queryIdx=220, trainIdx=340, imgIdx=0, distance=56.0] 
Good Matchs DMatch [queryIdx=221, trainIdx=72, imgIdx=0, distance=64.0] 
Good Matchs DMatch [queryIdx=222, trainIdx=278, imgIdx=0, distance=69.0] 
Good Matchs DMatch [queryIdx=223, trainIdx=221, imgIdx=0, distance=72.0] 
Good Matchs DMatch [queryIdx=224, trainIdx=339, imgIdx=0, distance=74.0] 
Good Matchs DMatch [queryIdx=225, trainIdx=155, imgIdx=0, distance=66.0] 
Good Matchs DMatch [queryIdx=226, trainIdx=278, imgIdx=0, distance=56.0] 
Good Matchs DMatch [queryIdx=227, trainIdx=165, imgIdx=0, distance=78.0] 
Good Matchs DMatch [queryIdx=228, trainIdx=279, imgIdx=0, distance=71.0] 
Good Matchs DMatch [queryIdx=229, trainIdx=355, imgIdx=0, distance=65.0] 
Good Matchs DMatch [queryIdx=231, trainIdx=69, imgIdx=0, distance=80.0] 
Good Matchs DMatch [queryIdx=232, trainIdx=278, imgIdx=0, distance=65.0] 
Good Matchs DMatch [queryIdx=233, trainIdx=65, imgIdx=0, distance=71.0] 
Good Matchs DMatch [queryIdx=234, trainIdx=93, imgIdx=0, distance=79.0] 
Good Matchs DMatch [queryIdx=235, trainIdx=203, imgIdx=0, distance=78.0] 
Good Matchs DMatch [queryIdx=236, trainIdx=159, imgIdx=0, distance=70.0] 
Good Matchs DMatch [queryIdx=237, trainIdx=93, imgIdx=0, distance=45.0] 
Good Matchs DMatch [queryIdx=238, trainIdx=172, imgIdx=0, distance=58.0] 
Good Matchs DMatch [queryIdx=239, trainIdx=374, imgIdx=0, distance=67.0] 
Good Matchs DMatch [queryIdx=240, trainIdx=278, imgIdx=0, distance=59.0] 
Good Matchs DMatch [queryIdx=241, trainIdx=223, imgIdx=0, distance=55.0] 
Good Matchs DMatch [queryIdx=242, trainIdx=365, imgIdx=0, distance=58.0] 
Good Matchs DMatch [queryIdx=243, trainIdx=91, imgIdx=0, distance=62.0] 
Good Matchs DMatch [queryIdx=244, trainIdx=238, imgIdx=0, distance=57.0] 
Good Matchs DMatch [queryIdx=245, trainIdx=299, imgIdx=0, distance=62.0] 
Good Matchs DMatch [queryIdx=246, trainIdx=289, imgIdx=0, distance=63.0] 
Good Matchs DMatch [queryIdx=247, trainIdx=93, imgIdx=0, distance=41.0] 
Good Matchs DMatch [queryIdx=249, trainIdx=5, imgIdx=0, distance=56.0] 
Good Matchs DMatch [queryIdx=250, trainIdx=93, imgIdx=0, distance=53.0] 
Good Matchs DMatch [queryIdx=251, trainIdx=93, imgIdx=0, distance=34.0] 
Good Matchs DMatch [queryIdx=252, trainIdx=97, imgIdx=0, distance=34.0] 
Good Matchs DMatch [queryIdx=253, trainIdx=93, imgIdx=0, distance=37.0] 
Good Matchs DMatch [queryIdx=254, trainIdx=174, imgIdx=0, distance=55.0] 
Good Matchs DMatch [queryIdx=255, trainIdx=91, imgIdx=0, distance=72.0] 
Good Matchs DMatch [queryIdx=256, trainIdx=81, imgIdx=0, distance=59.0] 
Good Matchs DMatch [queryIdx=257, trainIdx=92, imgIdx=0, distance=57.0] 
Good Matchs DMatch [queryIdx=258, trainIdx=212, imgIdx=0, distance=76.0] 
Good Matchs DMatch [queryIdx=259, trainIdx=119, imgIdx=0, distance=67.0] 
Good Matchs DMatch [queryIdx=260, trainIdx=228, imgIdx=0, distance=73.0] 
Good Matchs DMatch [queryIdx=261, trainIdx=119, imgIdx=0, distance=68.0] 
Good Matchs DMatch [queryIdx=263, trainIdx=266, imgIdx=0, distance=74.0] 
Good Matchs DMatch [queryIdx=264, trainIdx=319, imgIdx=0, distance=72.0] 
Good Matchs DMatch [queryIdx=265, trainIdx=157, imgIdx=0, distance=76.0] 
Good Matchs DMatch [queryIdx=266, trainIdx=365, imgIdx=0, distance=65.0] 
Good Matchs DMatch [queryIdx=267, trainIdx=341, imgIdx=0, distance=76.0] 
Good Matchs DMatch [queryIdx=268, trainIdx=303, imgIdx=0, distance=63.0] 
Good Matchs DMatch [queryIdx=269, trainIdx=313, imgIdx=0, distance=69.0] 
Good Matchs DMatch [queryIdx=271, trainIdx=350, imgIdx=0, distance=62.0] 
Good Matchs DMatch [queryIdx=272, trainIdx=313, imgIdx=0, distance=68.0] 
Good Matchs DMatch [queryIdx=278, trainIdx=267, imgIdx=0, distance=69.0] 
Good Matchs DMatch [queryIdx=280, trainIdx=223, imgIdx=0, distance=71.0] 
Good Matchs DMatch [queryIdx=281, trainIdx=267, imgIdx=0, distance=71.0] 
Good Matchs DMatch [queryIdx=283, trainIdx=334, imgIdx=0, distance=72.0] 
Good Matchs DMatch [queryIdx=284, trainIdx=313, imgIdx=0, distance=63.0] 
Good Matchs DMatch [queryIdx=285, trainIdx=78, imgIdx=0, distance=76.0] 
Good Matchs DMatch [queryIdx=286, trainIdx=312, imgIdx=0, distance=73.0] 
Good Matchs DMatch [queryIdx=287, trainIdx=271, imgIdx=0, distance=68.0] 
Good Matchs DMatch [queryIdx=288, trainIdx=170, imgIdx=0, distance=64.0] 
Good Matchs DMatch [queryIdx=289, trainIdx=278, imgIdx=0, distance=64.0] 
Good Matchs DMatch [queryIdx=290, trainIdx=282, imgIdx=0, distance=70.0] 
Good Matchs DMatch [queryIdx=291, trainIdx=91, imgIdx=0, distance=65.0] 
Good Matchs DMatch [queryIdx=292, trainIdx=334, imgIdx=0, distance=67.0] 
Good Matchs DMatch [queryIdx=293, trainIdx=80, imgIdx=0, distance=72.0] 
Good Matchs DMatch [queryIdx=294, trainIdx=92, imgIdx=0, distance=47.0] 
Good Matchs DMatch [queryIdx=295, trainIdx=301, imgIdx=0, distance=44.0] 
Good Matchs DMatch [queryIdx=297, trainIdx=220, imgIdx=0, distance=78.0] 
Good Matchs DMatch [queryIdx=298, trainIdx=374, imgIdx=0, distance=76.0] 
Good Matchs DMatch [queryIdx=300, trainIdx=329, imgIdx=0, distance=74.0] 
Good Matchs DMatch [queryIdx=302, trainIdx=285, imgIdx=0, distance=77.0] 
Good Matchs DMatch [queryIdx=305, trainIdx=271, imgIdx=0, distance=80.0] 
Good Matchs DMatch [queryIdx=307, trainIdx=350, imgIdx=0, distance=76.0] 
Good Matchs DMatch [queryIdx=308, trainIdx=320, imgIdx=0, distance=71.0] 
Good Matchs DMatch [queryIdx=309, trainIdx=163, imgIdx=0, distance=76.0] 
Good Matchs DMatch [queryIdx=310, trainIdx=170, imgIdx=0, distance=73.0] 
Good Matchs DMatch [queryIdx=311, trainIdx=357, imgIdx=0, distance=65.0] 
Good Matchs DMatch [queryIdx=312, trainIdx=320, imgIdx=0, distance=62.0] 
Good Matchs DMatch [queryIdx=314, trainIdx=342, imgIdx=0, distance=75.0] 
Good Matchs DMatch [queryIdx=315, trainIdx=162, imgIdx=0, distance=72.0] 
Good Matchs DMatch [queryIdx=316, trainIdx=239, imgIdx=0, distance=74.0] 
Good Matchs DMatch [queryIdx=317, trainIdx=171, imgIdx=0, distance=56.0] 
Good Matchs DMatch [queryIdx=318, trainIdx=244, imgIdx=0, distance=61.0] 
Good Matchs DMatch [queryIdx=319, trainIdx=369, imgIdx=0, distance=77.0] 
Good Matchs DMatch [queryIdx=320, trainIdx=346, imgIdx=0, distance=67.0] 
Good Matchs DMatch [queryIdx=322, trainIdx=158, imgIdx=0, distance=78.0] 
Good Matchs DMatch [queryIdx=325, trainIdx=92, imgIdx=0, distance=73.0] 
Good Matchs DMatch [queryIdx=326, trainIdx=236, imgIdx=0, distance=76.0] 
Good Matchs DMatch [queryIdx=327, trainIdx=162, imgIdx=0, distance=70.0] 

Liczba dopasowań, które otrzymuję, jest taka sama dla tego samego obrazu, jak również dla różnych obrazów. Jestem naprawdę zdezorientowana? Można wyjaśnić w jaki sposób porównać dwa obrazy i stwierdzić, czy są one podobne lub nie za pomocą OpenCV

Here is somewhat że próbuję osiągnąć

Answer 1

Ponieważ używasz programu BRUTEFORCE matcher, zawsze uzyskasz najlepsze możliwe dopasowania dla wszystkich kluczowych deskryptorów z zapytania (szablonu) w pociągu (obraz zawierający zapytanie). tj .: BRUTEFORCE matcher zawsze znajdzie 100% dopasowania (najlepiej odpowiadające punkty kluczowe dla wszystkich kluczowych deskryptorów zapytania w deskryptorach pociągów).

Oznacza to, że należy filtrować mecze jako poprawne dopasowania (inliers) i niepoprawne dopasowania (wartości odstające).

Można to zrobić na dwa sposoby

kalkulacja 1.Distance

Korzystanie odległość, jak wspomniano przez Andrey Smorodov. Można użyć tej metody (Ale to nie zawsze zapewniają prawidłowe wyniki)

List<DMatch> matchesList = matches.toList(); 
Double max_dist = 0.0; 
Double min_dist = 100.0; 

for (int i = 0; i < matchesList.size(); i++) { 
    Double dist = (double) matchesList.get(i).distance; 
    if (dist < min_dist) 
     min_dist = dist; 
    if (dist > max_dist) 
    max_dist = dist; 
} 

LinkedList<DMatch> good_matches = new LinkedList<DMatch>(); 
for (int i = 0; i < matchesList.size(); i++) { 
    if (matchesList.get(i).distance <= (3 * min_dist)) // change the limit as you desire 
    good_matches.addLast(matchesList.get(i)); 
} 

2.Determine Maska

Można użyć findHomography dostać maskę, która pozwala odnaleźć inliers i odstających wyraźnie (Uznając uwzględnieniu kamery stwarzają perspektywę jest prawie poprawne)

 LinkedList<Point> objList = new LinkedList<Point>(); 
     LinkedList<Point> sceneList = new LinkedList<Point>(); 

     List<KeyPoint> keypoints_RefList = keypointsRef.toList(); 
     List<KeyPoint> keypoints_List = keypoints.toList(); 

     for (int i = 0; i < good_matches.size(); i++) { 
      objList.addLast(keypoints_RefList.get(good_matches.get(i).queryIdx).pt); 
      sceneList.addLast(keypoints_List.get(good_matches.get(i).trainIdx).pt); 
     }     

     MatOfPoint2f obj = new MatOfPoint2f(); 
     MatOfPoint2f scene = new MatOfPoint2f(); 

     obj.fromList(objList); 
     scene.fromList(sceneList); 

     Mat mask = new Mat(); 
     Mat hg = Calib3d.findHomography(obj, scene, 8, 10, mask); 

teraz maska ​​jest opcjonalne wyjście w findHomography który jest tablicą obu 1O r 0 wartość dla każdego meczu. Wartość maski dla odpowiedniego dopasowania to 1, jeśli jest to wartość inlier i 0, jeśli jest to wartość odstająca.

Możesz użyć tego jako kryterium, aby zdecydować, czy masz prawie 90% maski na 1, aby można było uzyskać zgodność z prawdą.

używam go w rozpoznawaniu konkretnych obiektów z android ramie kamery java i mam te wyniki dziennika

08-22 01:08:38.929: I/OCVSample::Activity(25799): Keypoints Size: 1x477 KeypointsRef Size : 1x165 
08-22 01:08:39.049: I/OCVSample::Activity(25799): descriptor Size: 32x477 descriptorRef Size : 32x165 
08-22 01:08:39.129: I/OCVSample::Activity(25799): Matches Size: 1x165 
08-22 01:08:39.129: I/OCVSample::Activity(25799): matchesList Size: 165 
08-22 01:08:39.139: I/OCVSample::Activity(25799): Max dist : 460.44110107421875 Min dist : 100.0 
08-22 01:08:39.139: I/OCVSample::Activity(25799): good matches size: 19 
08-22 01:08:39.139: I/OCVSample::Activity(25799): obj size : 1x165 
08-22 01:08:39.139: I/OCVSample::Activity(25799): scene size : 1x165 
08-22 01:08:40.239: I/OCVSample::Activity(25799): Homography mask size : 1x165 
08-22 01:08:40.239: I/OCVSample::Activity(25799): Homography mask : [1; 1; 1; 1; 1; 1; 1; 0; 1; 0; 1; 1; 1; 1; 0; 1; 1; 1; 0; 1; 1; 1; 1; 0; 1; 0; 1; 1; 1; 1; 1; 1; 0; 0; 1; 1; 1; 1; 0; 1; 0; 1; 1; 1; 1; 1; 0; 1; 1; 1; 1; 1; 1; 1; 0; 1; 1; 0; 1; 0; 1; 0; 0; 1; 0; 1; 1; 1; 1; 1; 1; 1; 0; 1; 1; 1; 1; 1; 0; 1; 1; 1; 1; 0; 0; 1; 1; 0; 1; 1; 1; 1; 0; 1; 0; 1; 1; 1; 1; 1; 1; 1; 1; 1; 1; 1; 0; 1; 1; 1; 1; 0; 0; 0; 1; 0; 1; 1; 1; 1; 0; 0; 1; 1; 1; 1; 1; 0; 1; 0; 1; 1; 0; 1; 0; 1; 1; 1; 1; 1; 1; 0; 1; 1; 0; 1; 1; 1; 1; 0; 1; 1; 1; 1; 1; 1; 1; 0; 1; 1; 0; 0; 1; 1] 

3. Kolejny prostsze podejście byłoby porównanie histogramu tych dwóch obrazów do tego można użyć compHist(); funkcja openCV, jak pokazano here , a także odnoszą się do dokumentów openCV.

Różne metody w porównaniu histogramu dają zakres wyjściowy od 0 do 1 lub wyższej wartości, to wyjście zależy od podobieństwa między histogramami. Ostrożne w niektórych metodach 1 to 100% pozytywnego dopasowania i 0 w innej metodzie. "W przypadku metody chi-kwadrat niski wynik oznacza lepszą zgodność niż najwyższy wynik, idealne dopasowanie to 0, a całkowite niedopasowanie jest nieograniczone (w zależności od wielkości histogramu)."

Pozostałe: - Dwa całkowicie różne obrazy mogą mieć dokładnie tę samą wartość histogramu.

Porady:

1.Now dotyczące knnMatch wystarczy użyć matcher.knnMatch(); i odpowiednie typy danych wyjściowych.

2.Also w

matcher.match(query, train, matches); 

zapytanie => deskryptory KeyPoint dla szablonu-EG. kulka i pociąg = deskryptor punktu końcowego dla obrazu, który zawiera w sobie tę samą kulę. Liczba deskryptorów zapytania jest mniejsza niż liczba deskryptorów pociągu. Upewnij się, że masz rację.

Powodzenia.

Answer 2

Matcher szukać tylko na najbliższy mecz keypoints wykresie. Aby uzyskać różnicę między pomiarami, musisz użyć (średniej) sumy (lub innej metryki) odległości.