The riveting quality of aircraft skin is an important index to influence the performance of aircraft skin. On the basis of analyzing the elements of the riveting quality inspection of the skin

a visual inspection system for the riveting quality of aircraft skin is built. By analyzing the characteristics of image segmentation method in visual inspection system

we use a method that combine improved genetic algorithm with OSTU method to segment the aircraft skin image and get good segmentation results

and based on the segmented images

this paper introduces the method of extracting rivet feature and the way to obtain the detecting elements.