An Efficient Strategy for Face Clustering use in Video Surveillance System
Mahim-Ul Asad, Rashed Mustafa, Mohammad Shahadat Hossain · 2019
The importance of knowledge discovery from data has been increased dramatically with the increase of data over the past few years. A small video file contains more information compared to text documents and other media files such as audio, images. For this reason, extracting useful information from video i.e. automated video surveillance system has become a hot research issue. The presence of human in different frame of a video is a common scenario. In the security based application, identification of the human from the videos is an important issue. Face pattern is the most widely used parameter to recognize a person. A system with the ability of gathering the information about the presence of the same person in different frame of a video is highly demanding. In this study we have proposed a clustering algorithm which is variation of Hierarchical Agglomerative Clustering algorithm that is able to cluster face image of human. It tries to cluster based on both similarities and dissimilarities. Our proposed algorithm performs better compared to traditional Hierarchical Agglomerative Clustering algorithm in terms of accuracy and time complexity.