Interest Points Image Detectors: Performance Evaluation

Most of computer vision applications employs interest points of the scene (images or frames) to under stand it. Therefore an accurate detection of the interest points is essential for many computer vision applications. The interest points must be invariant to rotation, zoom, blur, illumination and point of view changes. This paper presents a performance comparison of the most popular interest points detectors, such as Harris, Harris-Laplace, Laplacian of Gaussian and SIFT, in order to know which of them could be the most accurate. In the evaluation, the average variation of the interest points through those changes is compared. Frame sequences of several environments are used to perform the evaluations. The comparison results can be used for selection of an adequate detector and furthermore improvement of the performance of them.