一种时空域联合的机动视频目标精确跟踪方法

提出一种采用Bhattacharyya系数最大化并联合时空域信息的视频目标跟踪方法。时域通过卡尔曼滤波预测目标的运动信息，空域用Camshift算法精确匹配视频目标。由于运动目标机动性比较强，卡尔曼滤波预测的位置和真实位置存在较大的误差，容易导致下一步跟踪失败。采用基于Bhattacharyya系数的由粗到精的核匹配搜索方法，在卡尔曼滤波预测的位置基础上适当扩大搜索范围，通过Bhattacharyya系数最大化确定初始匹配窗口，再用Camshift算法精确匹配视频目标。实验证明该方法对机动快速运动目标具有很高的跟踪精度。

Precise spatial-temporal tracking method for maneuvering video objects

A video object tracking method combining the Bhattacharyya coefficients maximization with spatial-temporal information was proposed. The Kalman filter was used to predict the target movement information in the time domain, while in the space domain the target was precisely matched by using Camshift algorithm. Due to the strong maneuverability of the moving target, there will be a relatively big discrepancy between the predicted and true position, which will cause failure in tracking for the upcoming frame. To deal with this problem, a kernel matching approach based on Bhattacharyya coefficients was adopted in a way from rough to precise. The search scope was properly increased based on the position of the prediction, and the initial matching window was defined according to the Bhattacharyya coefficients maximization. Finally, the target was precisely matched by applying Camshift algorithm. The experimental results indicate that the proposed method is highly precise in tracking fast maneuvering moving target.