Real-Time Surveillance System for Analyzing Abnormal Behavior of Pedestrians

In spite of excellent performance deep learning-based computer vision algorithms, they are not suitable for real-time surveillance to detect abnormal behavior because very high computational complexity. this paper, we propose a system analysis in closed-circuit television (CCTV) environment by constructing an algorithm and optimized CCTV environment. The proposed method combines pedestrian detection tracking extract information real-time, detects behaviors such as intrusion, loitering, fall-down, violence. To analyze behavior, it first determines intrusion/loitering through the coordinates object then fall-down/violence based on pattern object. is evaluated using intelligent data set distributed Korea Internet Security Agency (KISA).