Locally anomaly detection in crowded scenes using Locality constrained Linear Coding

Abstract

The investigation surrounding recent Stockholm and New York terrorist attack enforced this research to emphasize on anomaly detection. This paper describes the main part of an ongoing study through anomaly detection and localization which aims to improve anomaly localization accuracy. The sparsity constraint used in most recent anomaly detection researches is replaced with Locality-constrained Linear Coding. Locality-constrained Linear Coding (LLC) reconstruction cost criterion is designed to detect anomalies that occur in video locally. Implementing this method, the obtained experimental results approves considerable improvement regarding localization.