When car is in the process of high-speed, the faults caused by tire pressure anomalies are the most worried and difficult to prevent for all drivers. It is one of the important reasons for the sudden traffic accidents. Tire pressure related to ride comfort, fuel consumption, tire wear and vehicle safety, so tire pressure’s monitoring has been one of the important research directions in the safe driving of vehicle. The anomalies of tire pressure made the vehicles appear to tilt and shorten the distance between rim centres and the ground. Using image technology can detect intuitively and fast. This paper measures the dynamic tire pressure based on image processing technology. The vehicles tires goal was extracted by three image differences. After morphological processing, it used three methods to make comprehensive analysis and judgement of whether the tire pressure is normal, to realize the automation and intelligence. This method had realized real-time non-contact detection for dynamic tire pressure, which has extended significance to popularize.