Video Processing Technology Of Infrared Dynamic Scene Simulation System

Infrared radiation sources were required in many practical applications of Infrared Simulation System. These radiation sources can be an independent point. It can also be movement of objectives. With the rapid development of IRFPA, hot detecting infrared guidance has been replaced by imaging guidance to achieve the objectives of the identification.To evaluate the performance of guidance system, huge amounts of field flight test are needed, but this is expensive and time-consuming, and experimental system can't be experienced in all occasions. So people look forward to simulate the entire process of infrared guidance system target tracking in the laboratory, that is hardware-in-the-loop simulation, it can work in the infrared band, simulating the dimentional and temporal change of infrared radiation objectives in the apple of guidance system's eye. A group of dynamic change images relative to the real goal and background were getde by the geekers, which are used to evaluate and test the system performance. This requires a set of infrared scene simulation, to simulate the infrared radiation characteristics of the targets and the background, making the guidance systems as in the real environment.The dynamic infrared scene projection technology based on DMD is known as the micromirror array projection system (MAPS), the core of the devices Digital Micromirror Device (DMD) was manufactured by Texas Instruments. From the physical principles speaking, infrared images were obtained by modulated the launch of infrared radiation, it has a high spatial resolution and high frame rate, no dead pixels and good uniformity of characteristics.In this paper, the development of infrared image generation technology were induced in detail, and highlighted the system structure and working principle of dynamic infrared target simulator based on the DMD, for the working characteristics of DMD chip, the video format conversion technology were in-depth studied. A new search algorithm was proposed in the paper. The algorithm employs two search patterns, which the first pattern is a diamond and the other is a square, and the best match point would be got by the above two search mode. Finally motion vector can be worked out, and its superiority was also proved through software simulation. On the basis of the algorithm, the video image de-interlacing algorithms and video decoding algorithms were improved in the paper, and the feasibility of the algorithm were also tested through simulation.In the image scaling area, this paper presents an improved directional interpolation algorithm. Edge detection is to detecte the edge of regional and flat areas first, then to interpolate them respectively. The phenomenon such as sawtooth and faintness have been eliminated, the image resampling edge was more effectively, and the overall results have been improved.