Research And Implementation On Automatic Approach For Radial Distortion Correction

Radial distortion is the positional deviation of image pixel in the radial direction to the distortion center, which makes deformation in the image. And radial distortion is the main distortion of the image, which has the biggest impact on the image. Radial distortion of the image not only affects the visual effect, but also increases the difficulty of those subsequent image processing algorithms, image tracking and recognition etc.This paper proposes a new algorithm for automatically correcting the image radial distortion by analyzing the research status at home and abroad. The main steps of the proposed algorithm include contour extraction from the image edges, arc extraction from the contours, arc filtering, estimation of image distortion parameter and correction of image distortion. Research and experiments have been done on each technical aspect of automatic correction algorithm. Hence the proposed algorithm for automatically correcting image radial distortion has good results by choosing the suitable edge detection algorithm, proposing a new arc extraction from edge contours and a new arc filtering algorithm. And compared to the same type of algorithm it has lower time cost.The main research work of this paper are as follows:(1) For the problem of arc extraction from edge contours, a new arc extraction method has been proposed. The characteristic of the arc from the image edge contours has been found by studying and analyzing the image edge contours and combining the characteristics of the arc. The new method can extract arcs from the image edge contours quickly and accurately.(2) For the problem of arc screening, a new are screening method has been proposed. Studying and analyzing the extracted arcs, arcs extracted from the distorted image can be divided into two categories:Good arcs and Bad arcs. Good arcs can help to correct image distortion, while Bad arcs will have a bad impact on the correction results. The common invariant of Good arcs is obtained by formula derivation in the new method, and Good arcs are selected using the common invariant.(3) For the problem of automatic and robust correcting image radial distortion, a new automatic approach for radial distortion correction has been proposed. This new method uses appropriate image edge contour extraction algorithm, proposes a new arc extraction algorithm from edge contours and a new arc screening algorithm, and achieves an automatic correcting process of the image radial distortion. And the experimental results show that new automatic approach for radial distortion correction has the following features:fast, effectiveness and reliability.