Research And Implementation Of The Camera Calibration Technology

In the production of life, the eye is an important organ of human to obtain visual information. And the eyes are the main route for human to perceive the world and understanding the world. With the rapid development of society and science technology, the demand for digital information is increasing. Because of a series of theories of signal processing and image processing proposed, the technology of computer vision has been developed rapidly. Stereo vision is one of the hot research directions in computer vision, and camera calibration is the basis and necessary part of stereo vision technology. Therefore, camera calibration is a hot research today, attracting much attention and support. It has become an important research topic for many researches, and it also has the important theoretical significance and practical value.Simply speaking, camera calibration is the process by computer image coordinates and world coordinates to solve the mathematical model of camera imaging process. So far, there are many methods of camera calibration, but each method has its own advantages and disadvantages. Aiming at the disadvantages of some methods, a method for camera calibration based on round signs point is proposed. The study dependents on the plate with circular markers for the calibration template, using Zhang ZY calibration algorithm and combining with digital image processing technology. The main contents are as follows:(l)The relevant theoretical knowledge and algorithm of camera calibration are learnt. Meanwhile, imaging model and calibration related to the coordinate transformation are analyzed and the calibration template hardware acquisition environment is established. In order to obtain the target image, these acquisition images are processed and analyzed. The process includes image preprocessing in gray image, binary, image edge detection, image denoising. The image denoising of the perimeter denoising, area denoising and circularity denoising are applied to calculate the centroid coordinates of feature points, and finally computer image coordinates are obtained.(2)Combined with the calibration algorithm of Zhang ZY, to determine the relationship between the three-dimensional coordinates of the calibration board round signs point and image points. The monocular and binocular camera parameters are calculated to achieve through MATLAB, and the distortion coefficient is improved. The results using maximum likelihood estimation is optimized. Analyzing calibration results and calibration residuals, the expected experimental results are obtained. This calibration method has been proved to be feasible and effective.