On the Robustness of a New Boundary-Based Corner Detection Algorithm

Abstract

Corners play an important role in computer vision, such as curve fitting, optical inspection, image segmentation, object recognition, etc. In our previous work, we proposed a new boundary-based corner detection algorithm. In this paper, we will compare it with other methods by conducting a rigorous experiment to evaluate the performances of both the curvature measurement and the optimization procedure. The experimental results are rather promising, they show that our new algorithm outperforms other algorithms.