At present, police have only a single means of evidence collection in the process of law enforcement and can neither analyze and deal with evidence effectively in real time nor provide effective support for investigation and arrest of law enforcement tasks. To solve these problems, we propose a single police law enforcement system based on rights protection law enforcement tasks. We adopt network architecture, integrate advanced forensic equipment and transmission equipment and apply deep learning algorithm to the evidence processing system to construct a systematic single police law enforcement system based on single police evidence collection and multiple police cooperative evidence collection. We use system engineering method to design the system and the operation process. In the design process, we put forward key technologies such as information integration, command assistant decision making and feature extraction.