Linear Programming for Matching in Human Body Gesture Recognition

We present a novel human body gesture recognition method using a linear programming based matching scheme. Instead of attempting to segment an object from the background, we develop a novel successive convexification linear programming method to locate the target by searching for the best matching region based on a graph template. The linear programming based matching scheme generates relatively dense matching patterns and thus presents a key feature for robust object matching and human body gesture recognition. By matching distance transformations of edge maps, the proposed scheme is able to match figures with large appearance changes. We further present gesture recognition methods based on the similarity of the exemplar with the matching target. Experiments show promising results for recognizing human body gestures in cluttered environments.

• Type: Conference
• Year: 2005
• Where: AMFG
• Authors: Hao Jiang, Ze-Nian Li, Mark S. Drew
• Keywords: AMFG 2005, Biometrics, Body Gesture Recognition, Gesture Recognition, Human Body Gestures