An Advanced Harris-Laplace Feature Detector with High Repeatability

An advanced Harris-Laplace is proposed to remove the redundant points detected by original Harris-Laplace. In this novel method, all points detected at each scale are tracked and grouped beginning with the largest scale in the scale-space to make each group represent one local structure firstly. Then the point in each group which simultaneously leads to the maxima of corner points measuring and scale normalization Laplace function is selected. Finally, these points are described and matched by scale invariant feature transform (SIFT) descriptor successfully. Experimental results indicate that the proposed method has higher repeatability than original Harris-Laplace.