In today's rapidly advancing technological era, intelligent parking lot recognition, positioning, and navigation systems have become an indispensable part of large public venues such as commercial complexes and hospitals. Traditional parking lots suf‐ fer from various issues including weak signals, low navigation accuracy, and difficulty in finding parking spaces. To enhance parking efficiency, we have designed a machine vision-based parking space detection and path planning system. This system utilizes ultra‐ sonic sensors to gather parking space information and employs YoloV5 image recognition technology to analyze the license plate in‐ formation of incoming vehicles. All data is transmitted to a backend server and then presented to customers through a WeChat miniprogram. Using advanced path planning algorithms, the system can identify the nearest available parking space to the customer and provide navigation services.