| Human-vision-based pedestrian recognition plays a vital role in the intelligent monitoring system. Application of this technology in intelligent monitoring system can detect moving people in complex videos. When unusual circumstances occur or someone special event is detected, the system will timely warn to effectively protect persons and property for safety, it also helps the public security system to find the whereabouts of the suspects. To achieve those objectives, this paper proposes a moving people recognition algorithm based on the visual color processing mechanism. In order to simulate the real surveillance systems, we make a pedestrian dataset from Multi-Camera Videos (called MCV dataset).The proposed pedestrian recognition algorithm is inspired by a mechanism of the rod cells and cone cells in human visual system for the surrounding things process. The algorithm is constructed of two parts. In the first part, after capturing the pedestrians’ images from the videos, a spiking neural network is proposed to extract the color features of the images, and a set of new low-dimensional features are generated by fusing the color features and color moments. In the second part, a Support Vector Machine is trained and then used to recognize a specific people after feature reduction. The algorithm has been successfully applied to recognize people in CASIA Database with a high recognition rate. Experimental results demonstrate that the pedestrian recognition algorithm is comparable to state-of-the-art approaches in terms of accuracy. Furthermore, in order to evaluate performance and analyze characteristics of people recognition algorithms in multi-camera scenes, Multi-Camera Video (MCV) dataset is made in this paper. It is used to evaluate and analyze the proposed method, and the advantages and disadvantages of the proposed mothed are revealed. The direction for further improvement of the proposed pedestrian recognition algorithm is provided. |
PDF Full Text Request