PERSON RE-IDENTIFICATION BASED ON VIDEO FACE FEATURES IN REAL TIME FRAMEWORK

One of the main tasks is to distinguish a person captured in a police work event, such as a face recognition, picture or video. This means matching the faces in each still pictures and video footage. Automated facial acknowledgment for pictures with principle quality can perform palatable execution, aside from video-based facial acknowledgment, which is hard to accomplish.Recognition compared to still images, many disadvantages of their area unit video footage. Facial picture varieties, for example, enlightenment, introduction, posture, impediment and movement, are halfway genuine in video scenes. During this undertaking, we will actualize a video coordinating way to deal with coordinate pictures by giving up recordings Gross man Multiple Learning Approach and Convolutional Neural Network formula to understand unknown competition. Finally, give voice alerts in real-time situations unknown. Combine SMS alert and email alert at unknown face detection.