Member Positioning Method Based on Vector Orientation Analysis

This paper introduces a member positioning method based on vector orientation analysis, aimed at enhancing the edge detection accuracy and stability of truss climbing robots in complex environments. Traditional edge detection algorithms like Canny and LSD, though effective in extracting lines from images, are sensitive to environmental noise and changes in illumination, leading to edge displacement and discontinuities that affect positioning accuracy. By integrating the Canny algorithm with the horizontal line angle from LSD and employing vector orientation analysis for precise edge extraction, this study overcomes the impact of environmental disturbances, significantly improving the system's adaptability and recognition accuracy. The research validates the effectiveness of this method under simulated real-world conditions through experimental environment construction, accurately identifying and restoring the edge information of members, thus providing technical support for the safe and stable operation of truss climbing robots. This work not only enhances the accuracy of edge detection but also offers a new solution for member recognition and positioning in complex visual environments.