Research on Geometry Contour Positioning Technology of Small Components Based on Improved Canny Algorithm

Machine vision is widely used in industrial robotic arm grasping scenarios. In the face of the actual complex industrial environment, the contours of the identified components are affected by the lighting system. In most cases, shadows are included, and the components are small in size. Under the premise of taking the beat, the industrial camera cannot capture the perfect target contour, and the quality of shadow removal directly affects the grasping accuracy of the visual robot, which will lead to grasping failure in severe cases. This paper proposes a shadow contour extraction technology for small components based on the improved canny algorithm. The threshold processing method is used to segment the gray histogram, the shadow area and the background area are fused, and the improved canny algorithm is used to detect the edge of the component image. Get the detailed features of the image. Adaptive threshold segmentation and connected domain analysis are performed on edge images to achieve precise positioning of geometric contours.