Silhouette based gait recognition based on the area features using both model free and model based approaches

Gait recognition has lately attracted the attention of a plethora of computer vision and biometric systems researchers. It offers advantages like non-invasiveness, data acquisition from a distance and low resolution. The geometric area between the limbs derived from the spatial-temporal silhouettes serves as an effective feature that is capable of differentiating subjects. Both model-based and model-free approaches are tried while extracting features for the gait recognition. Support Vector Machine (SVM) classifier is used for training the features from silhouettes of the known gaits and identifying the unknown gaits based on the same features. Experimentation on the silhouette samples of publicly available CASIA database vindicates the supremacy of the model based approach over the model free approach.