With the rapid development of China's economic construction, road traffic is playing an increasingly important role in national economic construction. The ensuing road pavement maintenance and management problems highlight increasingly, among them the damage detection of road pavement become one of the priorities of relevant road maintenance department. In addition, a number of early-constructed high-grade highways along enter long or overhaul period, in order to improve the service life of the road, obtain pavement damage data has become the important topic of pavement maintenance management department. At present, the way of road pavement detection still is mainly artificial detection, but the artificial detection method is of low detection efficiency, timeconsuming and affects the normal traffic. Therefore, the study of pavement damage detection system has certain practical significance. This paper mainly expounds the research of pavement crack detection based on image processing technology. This paper first expounds the relevant pavement damage detection system. In this paper, the digital image processing methods are studied, including image enhancement processing, image edge detection and image segmentation technology. Digital image processing is the research emphasis of this article, and at the same time, it is also the core content of pavement crack detection technology. Finally, the road pavement crack image classification recognition is studied and several common methods of pattern recognition are introduced.