Research Of Automatic Exposure And Autofocus Image Processing Algorithm For Visible Light Camera

Automatic exposure and Autofocus are key technologies in the process ofvisible light camera, and it is an importment part to ensure image quality. AE and AFmethod based on image processing have a high degree of intelligent, and it has beenwidely used in the ordinary digital camera. This article in view of the airbonecamera’s imaging characteristic and problem, reasch and analysis AE and AFalgorithm based on image process.First of all, the theory of automatic exposure (AE) of visible light camera isexplained. Acoording to the exposure function of the camera, this dissertationbuilded the mathematical models of two successibe frames image brightness andexposure parameters. The airbone camera may face kinds of different scenes and thescene changes every time when it is working. An effectual automatic exposurealglrithm based on image histogram and means of block image is proposed in thispaper, it needs to analyze luminance distributing of one frame iamge and account theoptimal mean of the next image. An automatic exposure mode is designed for onecamera via the anlysis of the influence of exposure time and gain. Simulationexperiment and the actual camera application show that images obtained by theproposed algorithm have appropriate brightness. And it not only has a goodcapability to adapt to the different scene, but also it can meet the requirements of therun-time.Secondly, the airborne camera will defocus due to the the change ofenvironmental factors such as temperature and pressure when it is working in thehigh shy. This dissertation introduces an autofocus mode with temperature foucs anddistance compensation for airborone, and analyzed the theoretical feasibility. Animage quality assessment function using image edge sharpness proposed for rasterimage. The experiment compared with classical evalution function shows that it hasa smaller amount of calculation, better antinoise performance and higher precisionetc. Combining the application of camera, we designed an adaptive step-sizehill-climbing search strategy. We set up auotfocus algorithm simulation platformwith a PC computer, the experiments show that this mothed will accurately and fastfocus.Finally, considering the real-time requirement of automatic exposure and autofocus, this paper proposes an AE and AF system based on the FPGA embeddedsystem. The algorithm partitioned hardware and software parts. Software part isimplemented with C language in core PowerPC and hardware part utilizes FPGAlogic resources. We mainly studied the realization and optimization of the divider,histogram statistics, image acquisition and median filter in logic circuits. When thesystem clock is60MHz, the automatic exposure IPcore needs six microseconds,automatic focus needs about11.3times in the average adjustment and the camera isable to accurately focus, it meets the requirements of real-time.