Abstract: In recent years, with the development of economy, the coverage of power grid is increasing year by year, and the operating mileage of transmission lines is also increasing year by year. In order to avoid damage to transmission lines, causing large-scale blackout and causing serious economic losses, the WorldView-3 high-resolution satellite images were used as data sources to conduct inspection research on transmission channels based on multi-scale segmentation technology coupling feature optimization method. The research shows that using high-resolution satellite remote sensing data to carry out transmission channel inspection based on multi-scale segmentation technology coupling feature optimization method, the overall accuracy of inspection can reach over 86%, and the extraction accuracy of highway, water body and vegetation can reach over 95%. And according to different types of ground objects in the inspection of transmission channels, the inspection sequence of ground objects will have an important impact on the accuracy of experimental results. Then compared with the traditional manual interpretation and supervised classification methods, the method presented in this paper shows higher efficiency and applicability for transmission channel inspection based on high-resolution satellite remote sensing images.