Study On Camera Calibration And Distortion Correction Method Of Image

With the development of computer graphics, pattern recognition etc, the computer vision has been developed fast. Acting as human vision system with the computer and getting the geometrical information of the object in the three dimensional space from the pictures taken by the computer, so that rebuilding and recognizing object is final purpose of the computer vision study. The position of the space object and the relation with corresponding points in the object picture is dependent on the imaging model of the camera, the parameter in which is the camera parameter, and the process to resolve these camera parameters is called camera calibration. The camera calibration is wildly used in three dimensional measurements, three dimensional object rebuilding, object recognition, machine navigation, biomedicine and industrial inspection. The camera calibration has been one of main hot points in the scope of the computer vision.There are generally two kinds of methods for camera calibration: normal camera calibration methods and self-calibration. The basic method of normal camera calibration is based on the special physical experiment conditions, such as the form, size of the calibrated objects. The internal and external parameters of the camera can be resolved through image treatment. And the self-calibration method of camera is to resolve the basic matrix with the geometrical relation amount many different images, so the internal and external parameters can be resolved.With the influence of many factors, during image taking and transmission by the camera, the output image quality maybe not very high, and sometimes there will be distortion, so the image can not meet with requirements. In order to increase the image analysis result, these images must be corrected. The common correction methods are optical correction and digital correction. The digital image correction corrects the images taken by the camera with the mathematic analysis and image processing technology for the purpose of image analysis. With fast development of the digital image processing technology, it is prevalent for distortion correction with digital image correction methods.This paper briefly introduces the mathematic background and imaging geometry theory, describes the traditional methods on camera calibration, and then analyzes the affecting factors in the camera calibration. The camera calibration method which is used widely at present is discussed, and analyzes the robustness. Based on existent camera calibration theory, this paper puts forward a kind of fast camera calibration methods based on the corner point inspection. This paper also studied the digital image correction methods, put forward a kind of correction arithmetic which is better than normal method, and proved the convenience and efficiency by the experimental method.