The Research And Realization Of Distortion Correction Algorithm For Wide-angle Image

With the development of computer technology and all kinds of advanced intelligenttechnology, the camera was widely used in traffic, medical and other fields. In order toobtain a large field of scene information, the wide angle lens usually used whenshooting. However, using wide-angle lens will inevitably produce nonlinear distortion,which means the photographed image distorted and appeared the phenomenon ofdistance ratio. Although this distortion does not affect the sharpness of the image, butwill reduce image location accuracy, and then make it lost its application value in thefield of quantitative analysis and image measurement. In view of this situation, thispaper proposes an algorithm of distortion correction of wide-angle image.This paper mainly studies the wide-angle image distortion correction technology,and analyzes the existing distortion correction algorithm in detail. Then used MATLABsoftware to simulate these algorithms, researched the scope of application and thecorrection effect through a lot of experimental. On this basis, a new method of distortioncorrection was proposed, and observed the correction effect of distortion image inMATLAB. The algorithm used the method of grid template correction, according to themapping relationship between the distortion and ideal image, and get the offset surfacein the X axis and Y axis about distorted pixels. Then using the three B interpolationfunction to interpolate the surface, which gets the offset surface distortion of pixel.According to offset surface and coordinate distortion point will realize coordinatetransformation of wide-angle image. Finally, by bilinear interpolation method toreconstruct the undistorted image gray level, so as to realize the correction ofwide-angle image.The algorithm, depending on the solid theoretical foundation, used the MATLABsoftware to analyze the feature distortion image. And continuously optimize algorithmthrough a lot of simulation experiments. The algorithm was transplanted to DSP, on thecombination of soft and hard cases, verified from the overall applicability, practicability,timeliness, accuracy and other aspects, fully meet the design ideas, from the theoreticalstage to application stage more step. The experimental results show that the algorithmcan quickly and efficiently in wide-angle distortion image correction.