SYSTEM FOR PEDESTRIAN DETECTION AND ATTRIBUTE EXTRACTION BASED ON A JOINT DEEP NETWORK

摘要

The present invention relates to a system for pedestrian recognition and attribute extraction based on a combined deep layer network, which learns an image inputted from a surveillance camera through an artificial intelligence technique; extracts partial features of each body part including a face of the pedestrian and attributes of the pedestrian such as clothes, gender, body shape, a possessed item, etc.; and accurately figures out whether the pedestrian included in the image is the same as a certain pedestrian by using the extracted partial features and the attributes of the pedestrian, thereby effectively tracing a moving path of the certain pedestrian.