Human Action Recognition: Contour-Based and Silhouette-Based Approaches

摘要: Human action recognition in videos is a desired field computer vision applications since it can be applied human interaction, surveillance monitors, robot vision, etc. Two approaches of features are investigated this chapter. First approach contour-based type. Four such as Cartesian Coordinate Features (CCF), Fourier Descriptors (FDF), Centroid-Distance (CDF), and Chord-Length (CLF). The second silhouette-based Three Histogram Oriented Gradients (HOG), Optical Flow (HOOF), Structural Similarity Index Measure (SSIM) features. All these simple to compute, efficient classify, fast calculate. Therefore, demonstrate promising for recognition. Moreover, the classification achieved using two classifiers: K-Nearest-Neighbor (KNN) Support Vector Machine (SVM). experimental results demonstrated that have potential useful videos.