Design of Monitoring System Based on Event Camera

To solve the problems of motion blurring and the small dynamic range of the optical monitoring system, this paper proposes a new type of intelligent monitoring system based on a new sensor-event camera. The system firstly preprocesses and visualizes the data collected by the event camera; secondly, it extracts the two-way directional gradient histogram feature of the event image, and then maximizes the classification of the positive and negative samples through the support vector machine to obtain feature detectors; Finally, it detects the reconstructed image of the event camera and automatically recognizes the target. The experimental results show that the method of using directional gradient histogram features combined with classifiers can obtain a higher detection rate for event camera image recognition. The system has low power consumption and can work in a larger illumination range without being affected by target movement, and has a good application prospect.