CAJ | 학술논문

为解决目标追踪中目标被遮挡的问题，文中设计了一种基于卡拉曼滤波的均值漂移目标跟踪算法。在选定目标的情况下，利用均值漂移算法迭代求得目标的位置，当目标被遮挡时，根据前一时刻目标的位置和速度等信息采用卡尔曼滤波预测，得到目标的位置。实验结果表明，该算法在有遮挡的情况下，准确性和实用性更强。
위해결목표추종중목표피차당적문제，문중설계료일충기우잡랍만려파적균치표이목표근종산법。재선정목표적정황하，이용균치표이산법질대구득목표적위치，당목표피차당시，근거전일시각목표적위치화속도등신식채용잡이만려파예측，득도목표적위치。실험결과표명，해산법재유차당적정황하，준학성화실용성경강。
Mean shift object tracking algorithm based on Kalman filtering is designed to solve the object occlu -sion problem in the object tracking field .The target location is obtained by mean shift algorithm for the selected tar-gets, and by Kalman filtering forecasting according to the location and speed of the previous time for the blocked tar -gets.Experimental results show that designed algorithm is more accuracy and practical when the target is blocked .