Detecting People in Cluttered Indoor Scenes
Motion is one of the important visual cues for scene analysis. It is particularly useful when the scene is clutter, such as in typical home or office environments. We present a motion segmentation algorithm that makes use of tempo-ral differencing to detect moving people in cluttered indoor scenes. The algorithm is devised based on a couple of per-ceptual organization principles. To deal with missing data, noise and outliers, a robust segmentation and grouping technique called tensor voting is employed. The resulting real-time people detector can handle the presence of multi-ple persons, and varying body sizes and poses. It requires no initialization,uses subjective threshold, which defines the minimum saliency of "significant" motion, and the only two parameters are scales (sizes) of the local neighborhood for region and contour analysis.
Paper in PDF format