【作者】 熊建平；

【导师】 吴建华； 刘守义；

【作者基本信息】 南昌大学 ， 通信与信息系统， 2005， 硕士

【摘要】 计算机视觉作为当今最为活跃而又富有挑战意义的研究领域,其研究内容和应用领域相当广泛。本文以建筑陶瓷生产线在线实时检测控制系统的开发为应用背景,对边缘检测技术、摄像机传统标定技术、自标定技术等相关技术进行深入研究。 计算机立体视觉中,首先必须解决的是三维物点与二维像点的对应关系问题。因此,摄像机标定是计算机视觉实现的前提和基本问题。摄像机标定主要分为传统的标定算法和自标定算法。由于摄像机镜头存在着畸变,对于要求精度定位的应用,需要进行畸变校正。传统的标定算法中各种算法存在着计算量大的不足,本文将利用Tsai的两步标定法,考虑一阶径向畸变的摄像机标定算法,它与其它优化算法相比具有简单实用、计算量小的特点。实验结果表明,该方法可以达到较高的精度。 论文首先在绪论中对边缘检测、计算机视觉技术做了简要地介绍。在第二章中,给出了图像处理的基本概念,研究和分析了图像处理的常用方法,这些方法包括图像灰度变换、图像平滑,图像锐化、边缘检测、二值化等等。第三章中,针对机器视觉系统的摄像机标定技术,介绍了几种常用的标定方法。在此基础上,提出了一种新的简单、快速的Tsai两步标定法。第四章对以前的工作进行总结并介绍今后的研究方向。更多还原

【Abstract】 Computer vision is an active and challenging research field, its research and application areas spread widely. Based on development of on-line and real time detection control system about ceramics, this dissertation is devoted to investigating some related techniques in computer vision, including edge detection,conventional calibration, self-calibration and so on.In computer stereo vision, the primary problem to be solved is the relations between the 3D points and 2D image points. So the camera calibration is the premise and the basic problem. There are conventional calibration and self-calibration algorithms in camera calibration techniques. Since the camera lens used in computer vision sustain a lot of nonlinear distortion, recent research efforts has been concentrated on the distortion correcting techniques. The optimization algorithms in conventional calibration field have shortcomings that the computing quantity is colossal. In order to improve it, Tsai two steps’ s algorithm was developed in conventional camera calibration in this dissertation, and a calibration algorithm considering radial distortion. Compared with the ordinary optimization algorithm of calibration, this algorithm is simplier,less computational and of high accuracy.In the first chapter, we introduce briefly in computer vision and edge detection. And then the paper puts forward the concept of image processing,studies and analyzes ordinary methods about image processing, which include image smoothing, edge detection, binarization and so on. Some calibration methods are recommended in the third chapter. On the basis of these methods, an easy method for camera calibration and correction is studied by the mathematical model of camera by use of projective theory. In the fourth chapter, we give conclusions and some perspectives.更多还原