Research On Correcting Image Distortion Of Aerial CCD Camera Based On FPGA

At present, the aerial imaging equipment plays a significant role in variousfields, especially in the aeronautical reconnaissance. When the aerial imagingdetection system shoots the target scene on the ground, we use the wide-angle lens toget a larger field of view for the shooting range as large as possible. However, adistortion problem would appear when the large field of view angle lens provides awide of imaging. At the same time, the distortion also gradually increases with theincrease of angle of view, which directly affects the geometry accuracy of imagingand subsequent image mosaic. So, the job for correcting the lens distortion must becarried out. Both the optical design and image processing methods can resolve thisquestion. The optical design method will affect the system aberration and the imagequality in the design at the same time of reducing distortion, further more, a largespace will be occupied and not be conducive to miniaturization requirement to aerialimaging equipment. It is difficult for correcting the high quality lens distortion,either from design, processing, manufacturing, cost, or fitness. While the imageprocessing with has the characteristics of highly integration, low power, intelligentand great speed can correct distortion image in high real-time. So, the imageprocessing was accepted for image distortion correction in this paper. In this paper, the causes of the airborne CCD camera distortion were analyzedin depth. Because of design of the camera itself, manufacture, install, environmentalcontrol and so on, distortion comes into being. With the field angle increases, thedistortion is more severe. In the meantime time, a detailed investigation and study ondevelopment of aerial camera distortion correction at domestic and abroad had beendone. This paper has made an in-depth analysis of some classic algorithm forcorrecting distortion and has laid a theoretical foundation for subsequentexperimental study.What’s more, this paper analyzed the light structure of aerial imaging, themechanism of distortion, distortional model, distortion rate and so on. Not only theprinciple of camera calibration and calibration model but also the solution process ofthe camera internal, extrinsic and distortion coefficient were expounded in detail.This paper put forward a optimization method of distortion correction aiming at thelimitation and deficiency of the general method, based on the existing theory.This paper accomplished the process of image acquisition and calibration inMatlab software development environment based on the PC color image distortioncorrection experimental system. This paper used the Calib_Toolbox based on Matlabfor extracting of camera parameters and optimizing parameter, and explaining theerror which would be brought in the experiment. Then, we could accomplish thedistortion correction of color image after subsequent programming processingincluding that we achieved geometric coordinate position correction and pixel grayinterpolation of the distortion image. Furthermore, a new and efficient backsteppingreconstruction pattern matching method for image distortion rate analysis wasproposed, which quantified the level of distortion.At last, this paper has designed and implemented a video images capture anddisplay with a VGA monitor system for analog camera based on FPGA. To meet thereal-time, a blue print of distortion correction algorithm block was designed and brought. This paper has achieved geometric coordinate position correction and pixelgray interpolation of the distortion image based on FPGA. Further more, the designis verified on the ModelSim simulation software. The experimental results showedthat the correction results of design based on FPGA and the simulation results inMatlab almost unanimously, and the speed could be achieved in real-timerequirements, which laid a good foundation for the practical application ofengineering.