Color Cast Correction Research Of Four-Band Image Based On Video Stream

In recent years, with the attention of the society to the field of security, the rapid development of video surveillance technologies have achieved great results. However, since the imaging component CCD/CMOS of the monitoring camera can receive not only the visible light but also the near-infrared light, there is a great difference between the collected image and the actual image. Currently, the IR-cut switch device is used to solve the influence of near-infrared light in the field of video surveillance. In the daytime, the IR-cut is switched to infrared cut filter(IRCF), and the near-infrared light is filtered out; while in the night, IR-cut is switched to the full lens to make full use of all the light. So no matter day or night, using the IR-CUT switch device can achieve the good imaging effect, but IR-cut is constantly switched, causing switching failure.In this paper, for the fixed camera video, firstly a fast moving object detection method based on the improved mixture of Gauss is proposed. Secondly, a four-band image color cast correction method based on the video stream is proposed, and it is combined the moving object detection and the four-band image color correction algorithm. Finally, the four-band image color cast correction system based on the video stream is implemented on the embedded platform. The experimental results show that the method can improve the correction efficiency of the four-band video color cast, and enhance the real-time performance of the video.