CAJ | 학술논문

针对固定搜索窗口均值漂移算法在目标运动速度过快或目标尺度发生变化而导致跟踪失败的问题，提出一种与卡尔曼滤波相结合的自适应窗口跟踪算法。首先用卡尔曼滤波算法对运动目标进行预测及更正，设定感兴趣区域，并利用均值漂移算法确定搜索窗口大小和位置，同时结合Bhattacharyya系数进行目标定位，实现视频中目标跟踪。通过对比分析和实验结果，改进算法对目标尺度发生变形时具有较好的鲁棒性和准确性。
침대고정수색창구균치표이산법재목표운동속도과쾌혹목표척도발생변화이도치근종실패적문제，제출일충여잡이만려파상결합적자괄응창구근종산법。수선용잡이만려파산법대운동목표진행예측급경정，설정감흥취구역，병이용균치표이산법학정수색창구대소화위치，동시결합Bhattacharyya계수진행목표정위，실현시빈중목표근종。통과대비분석화실험결과，개진산법대목표척도발생변형시구유교호적로봉성화준학성。
Mean shift algorithm with fixed searching window may fail in tracking due to the target moving too fast or the variation in target scale.Aiming at this problem,we propose a tracking algorithm with adaptive window which combines the Kalman filtering.First the Kalman filter algorithm is used to predict and correct the moving target,and to set the region of interest;then the mean shift algorithm is used to determine the size and position of search window,and at the same time the Bhattacharyya coefficient is used to locate the target and to realise the targets tracking in video.Through comparative analysis and experimental results,it is proved that the proposed algorithm has good robustness and accuracy on deformation of target scales.