A Tracklet-before-Clustering Initialization Strategy Based on Hierarchical KLT Tracklet Association for Coherent Motion Filtering Enhancement

Coherent motions depict the individuals’ collective movements in widely existing moving crowds physical, biological, and other systems. In recent years, similarity-based clustering algorithms, particularly Filtering (CF) approach, have accomplished wide-scale popularity acceptance field of coherent motion detection. this work, a tracklet-before-clustering initialization strategy is introduced to enhance Moreover, Hierarchical Tracklet Association (HTA) algorithm proposed address disconnected KLT tracklets problem input feature, thereby making proper trajectories repair optimize CF performance crowd clustering. The experimental results showed that method effective capable extracting significant patterns taken from scenes. Quantitative evaluation methods, such as Purity, Normalized Mutual Information Index (NMI), Rand (RI), F-measure (Fm), were conducted on real-world data using huge number video clips. This work has established key, initial step toward achieving rich pattern recognition.