Complex wavelet transform-based approach for human action recognition in video

Human action recognition is a challenging research in computer vision applications because variety of human actions can be misclassified as some other action types. In this paper, we proposed a method for human action recognition based on dual tree complex wavelet transform (DTCWT). DTCWT has better edge representation and approximate shift-invariant properties compared to real-valued wavelet transforms. Experiments are carried out on different standard action datasets including KTH and MSR. We have performed the proposed method on multiple levels of DTCWT. The proposed method is compared with other state-of-the-art methods in terms of different quantitative performance measures, and the results of the proposed method are found to have better recognition accuracy.