Research on Underwater Dynamic Target Recognition and Location Method Based on Binocular Vision
Aiming at the problem that the dynamic underwater target tracking and positioning process requires good real-time and robustness,an underwater dynamic target positioning method based on binocular vision is proposed.Fast guided filtering is used to extract the illumination component of underwater images,build the improved two-dimensional Gamma function,adjust its parameters by using the distribution characteristics of illumination components,and realize the adaptive image brightness correction under differ-ent underwater environments.Kalman filter is used to estimate the position of the target at the next moment,and taking the predic-tion space as the ROI region for image correction,it greatly reduces the running time of the algorithm.The target is masked in the HSV space,and it is accurately located by the binocular localization algorithm after recognition.Through the water tank testing,compared with multi-scale Gaussian function,bilateral filtering and other algorithms,the proposed method has a significant improve-ment in running speed,reaching 35 FPS,and it has a high positioning accuracy during the positioning process,with the average rela-tive errors of 3.59％,3.35％,1.42％in the X,Y and Z directions.The results show that the algorithm can meet the real-time and robustness requirements in underwater dynamic target tracking and positioning.
万方数据
维普
