Spatio-temporal Feature Extraction/Recognition in Videos Based on Energy Optimization.

Videos are spatio-temporally rich in static to dynamic objects/scenes, sparse to dense, and periodic to non-periodic motions. Particularly, dynamic texture (DT) exhibits complex appearance and motion changes that remain challenge to deal with. This paper presents an energy optimization method for feature extraction and recognition in videos. For noise and background jitter, Tikhonov regularization (TR) with eigen-vector and Frenet-Serret formula based energy constraints is also proposed. Different periodicity of DT can be adapted by the time-varying number of learning temporal frames. Optimal duration of an image sequence is determined from the temporal property of its eigen-values. Unlike state-of-the-art recognition methods, i.e., sparse coding and slow feature analysis, the proposed method can capture physical property of objects and scenes: velocity, acceleration, and orientation. Also, static and dynamic image regions can be locally classified. Owing to these spatio-temporal features, stability, robustness, and accuracy of feature extraction and recognition are enhanced. Using DT videos, the superiority of the proposed method to state-ofthe- art recognition methods is experimentally shown.