Search code examples
opencvimage-processingfeature-detectionsurfflann

OpenCV error in flann knnmatch

I'm writing an image matching algorithm based on surf. I initialize surf with 400 hessian threshold:

surf = cv2.SURF(400)

and did flann matching.

But it is running too slow (on large sets of images). Then I decided to do light testing initially, followed by deep testing. So I changed the hessian threshold value to 1200.

surf = cv2.SURF(1200)

It is much faster now.

matches = flann.knnMatch(des1,des2,k=2)

But on some images it throws the following error:

    OpenCV Error: Assertion failed ((globalDescIdx>=0) && (globalDescIdx < size())) in getLocalIdx, file /tmp/opencv-2.4.10/modules/features2d/src/matchers.cpp, line 163
Exception in thread Thread-1:
Traceback (most recent call last):
  File "/usr/lib/python2.7/threading.py", line 810, in __bootstrap_inner
    self.run()
  File "/usr/lib/python2.7/threading.py", line 763, in run
    self.__target(*self.__args, **self.__kwargs)
  File "indexThreadsDeep.py", line 97, in threadRun
    if(imageMatch(qDes,tDes)):
  File "indexThreadsDeep.py", line 69, in imageMatch
    matches = flann.knnMatch(des1,des2,k=2)
error: /tmp/opencv-2.4.10/modules/features2d/src/matchers.cpp:163: error: (-215) (globalDescIdx>=0) && (globalDescIdx < size()) in function getLocalIdx
Solution

  • The issue was coming because due to high hession threshold there is only one keypoint in some images.But in knnMatch i was looking for 2 nearest neighbours.