Local Detection of Occlusion Boundaries in Video

Occlusion boundaries are notoriously difficult for many pat ch-based computer vision algorithms, but they also provide potentially useful information about scene structure and shape. Using short video clips, we pres nt a novel method for scoring the degree to which edges exhibit occlusi on. We first utilize a spatio-temporal edge detector which estimates ed ge strength, orientation, and normal motion. By then extracting patches from e ither side of each detected (possibly moving) edglet, we can estimate and compare motion to determine if occlusion is present. This completely l ocal, bottom-up approach is intended to provide powerful low-level informa tion for use by higher-level reasoning methods.