Function compatibility with flann in opencv

Question

I am working on an image search project for which I defined / extracted key point functions using my own algorithm. Initially, I only extracted one function and tried to match using cv2.FlannBasedMatcher (), and it worked fine, and I implemented it like below:

Here vec is 2-d list of float values of shape (10, ) Ex: [[0.80000000000000004, 0.69999999999999996, 0.59999999999999998, 0.44444444444444448, 0.25, 0.0, 0.5, 2.0, 0, 2.9999999999999996] [2.25, 2.666666666666667, 3.4999999999999996, 0, 2.5, 1.0, 0.5, 0.37499999999999994, 0.20000000000000001, 0.10000000000000001] [2.25, 2.666666666666667, 3.4999999999999996, 0, 2.5, 1.0, 0.5, 0.37499999999999994, 0.20000000000000001, 0.10000000000000001] [2.25, 2.666666666666667, 3.4999999999999996, 0, 2.5, 1.0, 0.5, 0.37499999999999994, 0.20000000000000001, 0.10000000000000001]] vec1 = extractFeature(img1) vec2 = extractFeature(img2) q1 = np.asarray(vec1, dtype=np.float32) q2 = np.asarray(vec2, dtype=np.float32) FLANN_INDEX_KDTREE = 0 index_params = dict(algorithm = FLANN_INDEX_KDTREE, trees = 5) search_params = dict(checks=50) # or pass empty dictionary flann = cv2.FlannBasedMatcher(index_params,search_params) matches = flann.knnMatch(q1,q2,k=2) 

But now I have one more function descriptor for each key point along with the previous, but different length. So now my function descriptor has this form:

 [[[0.80000000000000004, 0.69999999999999996, 0.59999999999999998, 0.44444444444444448, 0.25, 0.0, 0.5, 2.0, 0, 2.9999999999999996],[2.06471330e-01, 1.59191645e-02, 9.17678759e-05, 1.32570314e-05, 4.58424252e-10, 1.66717250e-06,6.04810165e-11] [[2.25, 2.666666666666667, 3.4999999999999996, 0, 2.5, 1.0, 0.5, 0.37499999999999994, 0.20000000000000001, 0.10000000000000001],[ 2.06471330e-01, 1.59191645e-02, 9.17678759e-05, 1.32570314e-05, 4.58424252e-10, 1.66717250e-06, 6.04810165e-11], [[2.25, 2.666666666666667, 3.4999999999999996, 0, 2.5, 1.0, 0.5, 0.37499999999999994, 0.20000000000000001, 0.10000000000000001],[ 2.06471330e-01, 1.59191645e-02, 9.17678759e-05, 1.32570314e-05, 4.58424252e-10, 1.66717250e-06, 6.04810165e-11], [[2.25, 2.666666666666667, 3.4999999999999996, 0, 2.5, 1.0, 0.5, 0.37499999999999994, 0.20000000000000001, 0.10000000000000001],[ 2.06471330e-01, 1.59191645e-02, 9.17678759e-05, 1.32570314e-05, 4.58424252e-10, 1.66717250e-06, 6.04810165e-11]] 

Now, since each point function descriptor is a list of two lists (descriptors) with different lengths that is (10, 7,), so in this case I get the error:

setting an array element with a sequence.

when converting a function descriptor to a numpy array of float data type:

 q1 = np.asarray(vec1, dtype=np.float32) 

I understand that the reason for this error is the different lengths of the lists, so I wonder what would be the right way to implement the same?

Answer 1

You must define one descriptor of size 10+7=17 .

So the space descriptor is now 17, and you should be able to use cv2.FlannBasedMatcher .

Create a global descriptor of the correct size desc_glob = np.zeros((nb_pts,17)) and fill it in manually or find the Python method. Maybe np.reshape((nb_pts,17)) ?

Edit:

In order not to support one type of descriptor over another, you need to weigh or normalize the descriptors. This is the same principle as calculating the global descriptor distance from two descriptors:

dist (desc1, desc2) = dist (desc1a, desc2a) + lambda * dist (desc1b, desc2b)