Pfthb

I want to determine value of cards on on a bigger picture, having a template pic, for example.

My code below with ORB_create way.

img1 = cv2.imread('templates/j.png',0)
img2 = cv2.imread('templates/large j.png',0)
orb = cv2.ORB_create(edgeThreshold=8, patchSize=31, nlevels=8, 
fastThreshold=20, scaleFactor=1.4, 
WTA_K=2,scoreType=cv2.ORB_HARRIS_SCORE,
firstLevel=0, nfeatures=5000)

kp1, des1 = orb.detectAndCompute(img1,None)
kp2, des2 = orb.detectAndCompute(img2,None)

bf = cv2.BFMatcher(cv2.NORM_HAMMING, crossCheck=True)

matches = bf.match(des1,des2)
matches= sorted(matches, key=lambda x:x.distance)
print(len(matches))


img3 = cv2.drawMatches(img1,kp1,img2,kp2,matches[:10],None, flags=2)
cv2.imshow('s',img3)
cv2.waitKey(0)

Here are output.

At first case edgeThreshold=8 , at second edgeThreshold=12. As you can see, flags aren't correct! They compare white areas on the pictures, instead of 'J'. If I try to compare '9' or '7' with 'J' - result is wrong too. (there should not be matches, but they are). Maybe problem is in params? I changed 'edgeThreshold', 'scaleFactor' and 'nfeatures' - no sense.

That is even a good example of pictures, they may be angled or with broken prospect.

Also, I've tried AKAZE_create - there is no matches at all, zero.

 detector = cv2.AKAZE_create()
 (kps1, descs1) = detector.detectAndCompute(im1, None)
 (kps2, descs2) = detector.detectAndCompute(im2, None)

 #print("keypoints: , descriptors: ".format(len(kps1), descs1.shape))
 #print("keypoints: , descriptors: ".format(len(kps2), descs2.shape))

 bf = cv2.BFMatcher(cv2.NORM_MINMAX)
 matches = bf.knnMatch(descs1,descs2, k=2) # typo fixed
 print(len(matches))