This project will capture the video surveillance video stream through the video transmission protocol. It is based on the YoloV3 target detection algorithm based on the convolutional neural network in 2018. The human body recognition is performed in real time to determine whether there are suspicious persons in the monitoring target area. The Android app sends an anomaly signal and a live video to remind the user that a strange and suspicious person has entered the private area. Further, the alarm bell can be performed by the one-button alarm function of the mobile terminal APP.
Our team members are three students in the software engineering major of Nanjing University of Posts and Telecommunications.