一种基于归一化分割方法的集体行为识别算法研究

Research on A Collective Motion Recognition Algorithm Based on Normalized Cut

提出了一种基于归一化分割方法的集体行为识别算法,该算法通过运动个体的复杂特征定义个体间的相似性度量,得到集体行为聚类。实验结果表明,在多种视频场景中,该算法均能有效地识别集体行为模式;以真实聚类个数的差异作为评价指标,本方法具有更高的准确性。

In this paper, a collective motion recognition algorithm based on Normalized Cut is proposed to identify the collective behavior, research the collective behavior clustering through defining the similarity measure between the complex characteristics in individuals. The experimental results show that, the algorithm can identify the collective behavior patterns effectively in a variety of video scene, and comparing with existed meth- ods shows that this method has higher accuracy with true clustering number difference as evaluation index.