Detection And Tracking Of Space Debris In Starry Background

With the expansion of human space exploration,the problem of space debris is becoming more and more serious.Space debris is so damaging to all kinds of aircraft systems that it is urgent to detect and track space debris.Space debris is small and dim target in the image.Because of the less pixels,the low SNR ratio,and the lack of texture and shape information,the targets are difficult to detect and track.Based on the practical application and using the image processing technology and the ARM raspberry pie platform,a complete starry background dim target detection system is designed.The thesis is carried out in the following directions:(1)This thesis model and analyze the background of the star image,and the detection and elimination algorithm of the thermal pixel noise is improved.The background non-uniformity correction is carried out,and the method of centroid extraction and subdivision localization with high location accuracy is studied.(2)In view of the background movement problem of the star image,the joint correlation algorithm and the maximum common subsequence algorithm are compared,and the image registration method based on the LIS algorithm is improved.The algorithm has obvious advantages in the image rotation robustness and computing speed.The background difference method after registration is improved,which has high detection rate and low false alarm rate under low SNR condition.When the SNR is as low as 3.3(2.32<SNR<5.24),the detection rate can reach 96.5% and the false alarm rate is below 4%.(3)An algorithm based on the main star wheel coding is proposed,and the star background suppression target detection algorithm is studied based on this star pattern recognition algorithm.The algorithm improves the efficiency of the algorithm while reducing the probability of mismatch.The robustness of the algorithm is much better than that of the traditional grid algorithm and the Pyramid algorithm.The algorithm has the characteristics of short recognition time and small database capacity,and its recognition rate is better than 98%.(4)The small and dim target tracking algorithm of Kalman filter and particle filter is studied.The performance of the optimized algorithm is verified by real star image simulation,and the root mean square error is better than 1 pixel.Based on the raspberry pie hardware platform and the OpenCV vision library,the star space background debris detection and tracking system is designed,and the real-time performance of the algorithm is verified.