Adaptive Harris Corner Detection Algorithm Based on B-Spline Function

Harris corner detection is a classic corner detection algorithm, but using Gaussian smoothing filter, there is phenomenon of corner information missing and corner migration. In this paper, to have a data-fitting and low-pass characteristics of B-spline function introduced into the algorithm, construct a new filter, while introducing the idea of image sub-block and the excluding neighboring points to build a B-spline-based adaptive Harris corner detection algorithm. The new corner detection algorithm to extract corner uniformity, positioning accuracy, and noise suppression is good. Through the experiment, the new algorithm significantly improves the image corner detection performance.