Extracting corresponding-point on textureless object surface based on texture synthesis

Since the feature points which can be detected by the state-of-the-art intensity based point-detectors are always gather at the range with significant intensity change, such as textured portions or edges of an image, there is nearly no points in the areas of weak texture can be detected by traditional detectors. In this paper we describe a new technique for Interest points detecting and matching from weak texture stereo pairs. The detection and matching mechanism can be separated into three steps: firstly, the information on the large textureless areas well be enhance by adding textures through texture synthesis algorithm and affine transform; secondly, detecting the initial interest-point set by classical interest-point detectors; finally, graph cuts are used to find the globally optimal set of matching points on stereo pairs.