Edge Detection of Space Station Docking Mark Based on Improved Canny Operator

This paper proposes an improved algorithm combining mathematical morphology and Canny for edge detection of spacecraft docking marks to address the problems occurring during spacecraft near-field docking such as low accuracy of edge detection due to salt-and-pepper noise interference and easy generation of false edges. To eliminate the influence of ambient light in space, the author first converted RGB to CIE Lab color space to separate brightness and chromaticity information. The salt-and-pepper noise caused by space ionizing radiation and cosmic rays is filtered out by a combined filter. Then, the cropped docking mark image is subjected to OTSU binarization, and the high interference from noise and docking mark scale is further eliminated by a mathematical morphological low pass filter OCCO. Finally, edge detection is performed on the processed image with Canny. Experimental results show that the algorithm can effectively remove the interference from salt-and-pepper noise and scale while retaining the edge details of the image and reducing its false edges, and thus improve the accuracy and effectiveness of detection.