The Comparison and Application of Corner Detection Algorithms.
Jie Chen,Li-hui Zou,Juan Zhang,LiHua Dou
DOI: https://doi.org/10.4304/jmm.4.6.435-441
2009-01-01
Abstract:Corners in images represent a lot of important information. Extracting corners accurately is significant to image processing, which can reduce much of the calculations. In this paper, two widely used corner detection algorithms, SUSAN and Harris corner detection algorithms which are both based on intensity, were compared in stability, noise immunity and complexity quantificationally via stability factor η, anti-noise factor ρ and the runtime of each algorithm. It concluded that Harris corner detection algorithm was superior to SUSAN corner detection algorithm on the whole. Moreover, SUSAN and Harris detection algorithms were improved by selecting an adaptive gray difference threshold and by changing directional differentials, respectively, and compared using these three criterions. In addition, SUSAN and Harris corner detectors were applied to an image matching experiment. It was verified that the quantitative evaluations of the corner detection algorithms were valid through calculating match efficiency, defined as correct matching corner pairs dividing by matching time, which can reflect the performances of a corner detection algorithm comprehensively. Furthermore, the better corner detector was used into image mosaic experiment, and the result was satisfied. The work of this paper can provide a direction to the improvement and the utilization of these two corner detection algorithms.