High Precision Centroid Extraction Algorithm of Space Target for Autonomous Optical Navigation

In this paper, a high-precision centroid extraction algorithm for autonomous optical navigation of space targets is proposed. The target celestial body image is a light spot in the image plane. Firstly, the target celestial body image is segmented by threshold to realize the segmentation of background and light spots; Then the imaging characteristics are analyzed. If the light spot distribution is non-Gaussian, an improved gray centroid extraction algorithm is proposed to achieve sub-pixel centroid extraction. The region centroid extraction method is used to achieve preliminary centroid extraction, and then a sub-pixel correction method is performed based on gray value and image gradient to obtain a more accurate centroid position. Simulation results show that this method can achieve fast and accurate extraction of target centroid, and the extraction accuracy is less than 0.3 pixels, which can provide technical support for accurate optical navigation.