The present traffic monitoring system gathers a lot of data and uses algorithms to interpret images captured by cameras. These information are used to control traffic flow, identify license plates, and keep track of traffic accidents. Traditionally, the collected data are routed to data centers for processing and analysis before being delivered back to the operator. As a result, a lot of data going into the data center and traffic monitoring center limits network capacity and slows down the operator's ability to make decisions quickly. We suggested using the edge computing paradigm to address this issue by positioning edge nodes close to the cameras and doing data preprocessing on edge devices. As an edge device, the article considers the Raspberry Pi, as well as computer vision algorithms for pre-processing data at edge nodes and the process of transferring data to data centers with subsequent operational decision making by the operator. We believe that the use of the edge computing paradigm will help optimize network traffic and offload the computing power of the data center.