介绍了一种基于低成本CMOS摄像头的智能监控系统,实现了现场监控的实时化、智能化。系统可自动判别危险程度,可分别提供绿色、黄色和红色报警信息,并驱动相关附属设备工作。针对图像传输过程中的噪声,系统采用了对噪声图像使用局部算子的滤波技术来提高图像质量。根据对静态图像进行三原色分离后误差域的研究,达到背景与图像分离的目的,成功实现了图像的获取和处理。在低成本投入下,以简单的配置和简便的操作达到了高性能监控系统的功能。-Introduce a low-cost CMOS camera-based intelligent monitoring system, realize the scene in real-time monitoring and intelligent. The system can automatically determine the degree of risk, they can provide green, yellow and red alarm information, and drive the work-related ancillary equipment. For image transmission in the process of noise, system noise images of the use of local filtering operator technology to improve image quality. Based on static images RGB domain after the separation error research, background and images to achieve the purpose of separation, the successful implementation of the image acquisition and processing. Input in the low-cost, simple configuration and easy to operate to achieve a high-performance monitoring system.