The Research About Interior Orientation And Affine Distortion Correction Of Digital Imaging

Since the invention of the digital camera, digital photography has replaced the placeof the traditional shooting methods. However, in some specific areas (such as the field ofaviation), the traditional photographic film is still in use. It’s mentioned that affinedistortion will be generated in the process of photographic film being scanned anddigitized in some classic books about the discipline of mapping. So it’s needed todetermine the relationship between the pixel coordinates of the point in the ima ge and thephysical coordinates through the interior orientation. It’s a general affine transformationbetween the two coordinates. In the process of the interior orientation，parameters of themodel can be determined by the pixel coordinates and the corresponding physicalcoordinates of fiducial marks. But the relationship between pixel coordinates of points inthe image and the corresponding physical coordinates is described by the internalparameter matrix of camera in the discipline of computer vision, it is a special case ofaffine transformation. Obviously it’s not acceptable to take place of the general with thespecial. To make the description in the discipline of mapping consistent with thediscipline of computer vision, the affine distortion correction need to be carried on to thedigital image obtained by the scanning of photographic film. After the affine distortioncorrection of the image, a set of computer vision theory can be used. This reduces barriersbetween the traditional discipline of mapping and computer vision and makes the tiesbetween the two closer. Therefore it’s very necessary to carry on affine distortioncorrection to the digital image based on the context of interior orientation. The main workand innovations are as follows:(1) Based on the background of interior orientation, a method for correcting affinedistortion of digital image is proposed. According to the different description to the mathematical model which describes the relation between pixel coordinates and thecorresponding physical coordinates in the two disciplines of mapping and computervision, a method for correcting the affine distortion is proposed. The purpose of theproposed method is not only to correct the affine distortion, but also to make the contactbetween the mapping and computer vision more closely. This paper describes theprinciple of the method, and it’s validated through multiple sets of simulationexperiments.(2) A closed solution of the parameters of interior orientation is proposed. In thetraditional method, it’s needed to carry on the inversion of6th order matrix. The proposedmethod reduces the computational complexity by eliminating inverse operation of thematrix, as a result it improves the speed to solve parameters. This paper describes theprinciple of the method and a set of experiments are designed to verify it。...