search

danini / magsac

The MAGSAC algorithm for robust model fitting without using an inlier-outlier threshold
Other
415 stars 65 forks source link

Fail to run new version of pymagsac. #36

Open ACuOoOoO opened 3 months ago

ACuOoOoO commented 3 months ago

I follow README to install pymagsac from source. However, I cannot run the python examples. I have tried to fix bugs in the examples. However, no inlier cannot be found by

def verify_pymagsac_fundam(kps1, kps2, tentatives, use_magsac_plus_plus):
    src_pts = np.float64([ kps1[m.queryIdx].pt for m in tentatives ]).reshape(-1,2)
    dst_pts = np.float64([ kps2[m.trainIdx].pt for m in tentatives ]).reshape(-1,2)
    src_pts = np.ascontiguousarray(src_pts)
    dst_pts = np.ascontiguousarray(src_pts)
    H, mask,_ = pymagsac.findFundamentalMatrix(
        src_pts, 
        dst_pts,
        2592.0,1944.0,2592.0,1944.0,
        probabilities = np.ones(src_pts.shape[0]))
    print (deepcopy(mask).astype(np.float32).sum(), 'inliers found')
    return H, mask