Corner Detectors for Affine Invariant Salient Regions: Is Color Important?

Abstract

Recently, a lot of research has been done on the matching of images and their structures. Although the approaches are very different, most methods use some kind of point selection from which descriptors or a hierarchy are derived. We focus here on the methods that are related to the detection of points and regions that can be detected in an affine invariant way. Most of the previous research concentrated on intensity based methods. However, we show in this work that color information can make a significant contribution to feature detection and matching. Our color based detection algorithms detect the most distinctive features and the experiments suggest that to obtain optimal performance, a tradeoff should be made between invariance and distinctiveness by an appropriate weighting of the intensity and color information.

Semi-supervised learning of object categories from paired local features

“…Our work is close to [2,9] where matching is formulated as probabilistic relaxation. On the other hand, interest points have been actively studied and applied to image and video retrieval applications [20,27,16,22], which inspired our feature representation using only local keypoints in this task.…”

Local Invariant Feature Detectors: A Survey

“…Salient points are the maxima of the saliency map, which represents distinctiveness of color derivatives in a point neighborhood. In related work [217] they argue that the distinctiveness of color-based salient points is much higher than for the intensity ones. Color ratios between neighboring pixels are used to obtain derivatives independent of illumination, which results in color interest points that are more robust to illumination changes.…”

Interest points based on maximization of distinctiveness

“…A good overview is given in Sebe, et al [7] and also Schmid, et al [3,12]. In these works, it is clear that one of the best performing interest or salient point detectors is the Harris corner detector.…”

“…A distribution that is spread out, indicates more information content. Sebe, et al [7] suggest a good measure is using the information content as measured by the average information content of all messages or the entropy. Tian et al [6]l use the retrieval accuracy in a content based image retrieval task to evaluate their wavelet-based salient point method.…”