Analysis and Recognition of Walking Movements

We present an approach for recognizing human walking movements using low-level motion regularities and constraints. Biomechanical features for classification are automatically extracted from video sequences of walkers. A multiplicative classification rule using statistical distances is then used to determine whether an unknown motion is consistent with normal walking patterns. Recognition results are shown distinguishing walking examples across multiple speeds from other non-walking locomotions.

• Type: Conference
• Year: 2002
• Where: ICPR
• Authors: James W. Davis, Stephanie R. Taylor
• Keywords: Computer Vision, Human Walking Movements, ICPR 2002, Low-level Motion Regularities, Normal Walking Patterns