LSS-target Detection And Tracking Technology Based On Video Sequence

With the development of "low,slow and small" technologies such as UAV,the phenomenon of "black flying" has been banned repeatedly,and its negative impact on national security,social security and citizen privacy is becoming more and more serious,even hindering the normal development of the national economy.Based on the vehicle mounted "low slow small" prevention and control platform with laser gun and other directional energy weapons,this paper designs the optoelectronic tracking system architecture,and studies the key technologies in the optoelectronic tracking system,mainly including "low slow small" target detection,long-term stable tracking algorithm,automatic exposure parameter control,miss distance lag compensation.In the process of target detection,an adaptive parameter illumination equalization algorithm is proposed for the sky background with strong illumination,which effectively adjusts the brightness of the collected image and highlights the target.Then,the image is enhanced by combining Gaussian filter denoising and morphological transformation to suppress the background.Finally,the adaptive threshold is used to extract the "low slow small" target in the sky background,and the detection results are analyzed Output to follow-up tracking module.In the aspect of target tracking,aiming at the target characteristics of "low slow small" aircraft,as well as the tracking difficulties such as motion blur,low resolution,vulnerable to occlusion in practical application scenes,a long-term target tracking algorithm is proposed based on fast discriminant scale space tracking algorithm(f DSST),which uses confidence to update and occlusion judgment,and adds multi feature fusion and re detection module The experimental results show that the algorithm can achieve long-term stable tracking of "low slow small" targets,and the tracking accuracy and success rate are significantly improved compared with the benchmark algorithm.Aiming at the influence of "overexposure" phenomenon on miss distance extraction in the actual application scene of the prevention and control platform,an automatic exposure control method combined with normalized average gray value is proposed.According to the normalized average gray value,the image quality evaluation standard is optimized,and the exposure parameter adjustment value is determined through image brightness block statistics and related experiments.The strategy of "exposure time first,gain later" is adopted by adjusting the exposure time and image gain dynamically,the image quality and operation speed are improved.In addition,aiming at the problem that the time delay caused by the image processing system and the automatic exposure control affects the tracking algorithm and servo control accuracy,the time lag feed-forward compensation link is designed.Considering the characteristics of "low slow small" target,the parameter dynamic adjustment is proposed,the "current" motion model is improved,and the target information is predicted by combining with the extended Kalman filter,which improves the tracking performance of the platform Tracking accuracy,effectively compensate the delay of each link.