Print Email Facebook Twitter Single person pose recognition and tracking Title Single person pose recognition and tracking Author Barbadillo Amor, J. Contributor Hendriks, E.A. (mentor) Huo, F. (mentor) Faculty Electrical Engineering, Mathematics and Computer Science Department Mediamatics Programme Pattern Recognition and Bioinformatics Group Date 2010-06-25 Abstract The goal of this research is to improve a system capable to detect, track a single person and recognize poses real time for controlling a spatial game. After performing background subtraction, the human blob is segmented in order to track the torso and hands. Angles and distances between hands and torso center are used to compute the features. Finally, a 10-Nearest-Neighbor classifier recognizes 9 predefined Poses which are used by the player to control the game. This work contributes with two improvements. The first one is a more robust improved hand detection combining the current skin color detection with human blob information. The second improvement is a classifier that recognizes Non-Poses in addition to the 9 predefined Poses that are used in the game. Subject pose recognitioncomputer visionhuman action recognition To reference this document use: http://resolver.tudelft.nl/uuid:944a11fd-eea8-445e-9a46-98855f7766d3 Part of collection Student theses Document type master thesis Rights (c) 2010 Barbadillo Amor, J. Files PDF Single_person_pose_detect ... acking.pdf 1.49 MB Close viewer /islandora/object/uuid:944a11fd-eea8-445e-9a46-98855f7766d3/datastream/OBJ/view