Moving Object Detection In Video Streaming Using Improved DNN Algorithm

Nowadays security issues are becoming more difficult, video recognition is generally very difficult. In this paper, video detection aims to detect whether dynamic or static. We applied CNN to the tasks of video detection and recognition through image processing techniques. We constructed a dataset of the most frequent video items in Google, and used it to evaluate recognition performance. The network classifies the train data and test data and gives the results of classified output. However, deep learning has been shown recently to be a very powerful image recognition technique, and it is a state-of-the-art approach to deep learning. This network obtained of accuracy more than existing methods.