Abstract: A multi-type damage detection method for wind turbine blades is proposed by machine vision detection technology to promote the intellectual development of its operation and maintenance. Firstly, the blade image from the intelligent patrol UAV platform is used to identify the blade-damaged area by graying, filtering, enhancement, segmentation, and morphological processing. Then, an identification classifier of the blade damage type is designed through the geometric features and gray feature information of the blade damage area by connected domain analysis. Finally, the detection algorithm and classifier are integrated into the wind turbine blade damage visual detection system. The results show that the average detection accuracy is 90.4% for typical blode damage such as skin peeling, coating damage, sand holes, oil stains, cracks, etc.