Design and Research of a Bolt Hole Recognition and Positioning System Based on Machine Vision

This paper aims to explore a machine vision-based bolt hole recognition and positioning system. By analyzing and designing a system capable of precise identification and localization of bolt holes, it seeks to meet the demands of industrial automation assembly. In terms of three-dimensional recognition and positioning, both coarse and fine positioning methods are proposed. The coarse positioning section investigates the acquisition of the bolt's three-dimensional coordinates and its pose information, and implements a sorting method for bolts within a template. For fine positioning, an enhanced edge detection algorithm, Canny-OTSU, and a bolt hexagon detection algorithm, HT-AH, are employed. The system design section discusses hardware selection, including the choice of camera and supplementary lighting, and the accuracy of the positioning algorithm is demonstrated through experimentation. The research findings indicate that the designed system exhibits high precision and reliability in complex industrial environments, providing crucial technical support for efficient and automated assembly.