多线索融合的无视野重叠跨摄像机行人跟踪算法

针对单个摄像机视野有限而无法满足日益扩大的监控范围的现象，此文对无视野重叠的跨摄像机行人跟踪算法进行了研究，并提出了一种融合时空线索和外观线索的无视野重叠跨摄像机行人跟踪算法。文章在对已有摄像机网络拓扑结构估计算法分析的基础上提出了一种基于加权时间窗口的无视野重叠摄像机网络拓扑结构估计算法。然后利用朴素贝叶斯完成两种线索融合，实现不同摄像机间行人匹配和跟踪信息的传递，最终实现无视野重叠区域的跨摄像机行人跟踪。该算法在公开的MCT数据集上进行对比实验并取得了优于其它算法的结果。

Pedestrian tracking across non-overlapping cameras based on fusion multi-clue

For the phenomenon that the single camera view is limited and can not meet the growing scope of monitor, we study algorithms that pedestrian tracking across non-overlapping camera views and proposed a pedestrian tracking across non-overlapping camera views based on fusing spatial-temporal and appearance clues. On the basis of analyzing the existing camera network topology estimation algorithm, we proposes the cameras network topology estimation algorithm based on weighted time window. Then, we fuse two kinds of clues by using Naive Bayes and achieve pedestrian match and transfer of tracking information across different cameras. Lastly, we realize pedestrian tracking across non-overlapping camera views. Comparative tests in MCT dataset have been done, and results are better than other methods.