A Study Of Barrel Distortion Correction Based On RTCam Model

The space vision system is one of the most important equipment in space exploration. For each camera in the system, it needs to make up its own model. When many cameras need to work together in the image processing cooperation, each model needs to combine into a total one. Focusing on the question of unified modeling problems which are needed to be modified together, scientists from British have put forward a camera model which is called Rational-Tensor (briefly called as RTCam model), and established a model of variety types of cameras and some nonlinear constrains as well as the geometric framework.This paper uses RTCam model to correct the barrel distortion, uses an algorithm basing on RTCam model to correct the images with barrel distortion, and measures the correction accuracy by quantitative analysis. Firstly, this paper uses algorithm basing on RTCam model on the image collected by the normal lens without any barrel distortion to get the images with serious barrel distortion. then calculates the image with barrel distortion to get the parameter s, after above this paper uses the algorithm of barrel distortion model of RTCam to correct the barrel distortion image, by comparing the specific straight line of pre-distortion with post-distortion, the comparison of fitting error parameter illustrates that the fitting algorithm can turn the seriously barrel distortion correction curve into straight line, and makes a contrast coefficient and constant term of the linear fitting value description compared with no shift occurs in the image plane before and after the correction of the linear distortion, thus proves that the barrel distortion correction algorithm can be highly accurate images with serious barrel distortion.Then, uses an algorithm based on RTCam model to correct the images with barrel distortion collected by wide-angle lens. By comparing the specific straight linear fitting line of pre-distortion and post-distortion, the comparison of this algorithm which the tranditional algorithm quantitatively shows that this algorithm has a higher precision.