Transformer terminal feature recognition and positioning method based on binocular vision and image processing algorithm

Abstract

Based on binocular vision and image processing algorithm, the recognition and location of high voltage side and low voltage side of transformer were completed for transformer automatic wiring positioning platform. On the basis of binocular vision calibration and 3D point cloud reconstruction, P3P algorithm was used to complete the positioning of transformer operation space coordinate system. Based on the image threshold segmentation algorithm, the position of transformer high voltage terminal, low voltage terminal, and transformer rotation angle were identified on the basis of 3d point cloud data. The target position information was provided for transformer automatic wiring device.