Abstract
In the field of developing innovation, pictures are assuming as an important entity. Almost in all fields, picture base data is considered very beneficial, like in the field of security, facial acknowledgment, or therapeutic imaging, pictures make the existence simple for people. In this paper, an approach for both human detection and classification of single human activity recognition is proposed. We implement the pre-processing technique which is the fusion of the different methods. In the first step, we select the channel, apply the top hat filter, adjust the intensity values, and contrast stretching by threshold values applied to enhance the quality of the image. After pre-processing a weight-based segmentation approach is implemented to detect and compute the frame difference using cumulative mean. A hybrid feature extraction technique is used for the recognition of human action. The extracted features are fused based on serial-based fusion and later on fused features are utilized for classification. To validate the proposed algorithm 4 datasets as HOLLYWOOD, UCF101, HMDB51, and WEIZMANN are used for action recognition. The proposed technique performs better than the existing one.