Harris corner detection and localization in OpenCV with Python -

i'm using following code try detect corners of polylines in order 'measure' lines. code based on snippet found somewhere on so , based on cv2.cornerharris():

cornerimg = cv2.cornerharris( gray, # src                               2,    # blocksize                               3,    # ksize / aperture                               0.04  # k                                      # dst                                     # bordertype                               )  # ? cornerimg = cv2.normalize(  cornerimg,        # src                             none,             # dst                             0,                # alpha                             255,              # beta                             cv2.norm_minmax,  # norm type                             cv2.cv_32fc1,     # dtype                             none              # mask                             )  # ? cornerimg = cv2.convertscaleabs( cornerimg )  cornershow = cornerimg.copy()  # iterate on pixels corner positions w, h = gray.shape y in range(0, h):   x in range (0, w):     #harris = cv2.cv.get2d( cv2.cv.fromarray(cornerimg), y, x)     #if harris[0] > 10e-06:     if cornerimg[x,y] > 64:       print("corner @ ", x, y)       cv2.circle( cornershow,  # dest                   (x,y),      # pos                   4,          # radius                   (115,0,25)  # color                   )  cv2.imshow('harris cornerimg', cornershow) 

the original code results in white spots @ corner location , level seems indicator of "corneryness". snippet (updated use cv2) iterates on resulting image , scans values lager 10e-06 reason, have replaced comparison of think should brightness in image.

however, circles drawn @ locations near actual hot-spots found in normalized harris output.

what doing wrong?

alternatively, cv2.goodfeaturestotrack() can set use harris (useharrisdetector=true) attempt use not result in cornerharris appears detect properly:

cv2.goodfeaturestotrack(  blurred,                # img                                 500,                    # maxcorners                                 0.03,                   # qualitylevel                                 10,                     # mindistance                                 none,                   # corners,                                  none,                   # mask,                                  2,                      # blocksize,                                  useharrisdetector=true, # useharrisdetector,                                  k=0.04                  # k                                 ) 

what equivalent function call cv2.cornerharris()?

the output seems transposed, swapping x , y indices on square image fixes (circles on corner maxima).