Real-time Processor Of The Light Level Television

Low-light level television is an important branch of LLL night vision technology. It is a combination of television camera and monitor device, which works in a low light intensity environment. At present, LLL TV has been widely used in military, public security, astronomy, navigation, space industry and so on. Particularly, in the area of military, LLL TV plays an important role in night fighting, conducting, detecting, monitoring, tracking, aiming, fire control, nighttime driving and so on.In recent years, image fusion technology has been developed rapidly and become an important subject of night vision technology. Our laboratory has been engaged in the research and development of image fusion systems, which are mainly the fusion of infrared image and visible image, infrared image and LLL image. In the infrared image and LLL image fusion system, because of the inherent defects of LLL imaging device, the LLL image has the shortcomings of high random noise, narrow dynamic range of gray, low contrast and blurry edge. These shortcomings reduce the quality of LLL images seriously, thereby affect the output of fusion system. For this reason, the pretreatment of LLL images becomes critical, it is related to the quality of fusion image directly.This paper is committed to real-time processor of LLL TV, in order to obtain images with low noise, distinction and clear outline. This paper introduces the basic theory, current situation and development of LLL TV; fully analyzes the features of LLL images; researches the image quality evaluation methods, and then offers an evaluation method for LLL image, which combines both the subjective and objective; researches and analyzes the common denoising algorithms of LLL images, and then offers a self-adapting space-time filtering algorithm that based on quadratic motion detections; completes the solution that takes FPGA as the core for real-time LLL TV processor; finally, analyzes the results of real-time LLL TV processor.In this paper, the self-adapting space-time filtering algorithm based on quadratic motion detections has the advantages of better filtering capacity and lower motion blur. It has been achieved in the hardware platform of the real-time LLL TV processor, and got a relatively good output.