基于Harris角点的图像配准算法研究毕业论文.doc

东北石油大学本科生毕业设计（论文） 摘 要 图像配准现在已成为数字图像处理的研究热点，方法繁多，站在时代的前沿。图像配准多采用基于图像特征点的方法，这种方法易于用计算机处理并且容易实现人机交互，其重点在于如何提取图像上的有效特征点。 对图像拼接技术的目的、意义、国内外研究现状、发展方向以及本课题研究的目的和意义进行了阐述，着重介绍了图像拼接过程的核心技术——图像配准。阐明了现有配准方法的工作原理，并对常用的各种经典算法的优点和不足进行了比较总结。针对基于特征的角点检测算法，详细介绍了 Harris 角点检测算法。利用MATLAB7.0软件编程，实现图像的角点检测、图像配准以及图像拼接，并且设计精美的图形用户界面(GUI)，以演示最后的结果。 仿真结果证明：Harris算子具有较好的抗干扰能力，实现算法简单，易于编程，实现图像拼接具有较高的稳定性和鲁棒性，而且在图像存在灰度变化、旋转以及噪声等情况下，可以对特征点进行较好的提取，使角点检测的误检率低并且匹配效率高。 关键词：图像拼接技术；图像配准；Harris角点检测算法；MATLAB 7.0 Abstract Image registration is now a research focus in the digital image processing, has many kinds of methods, and keeping pace with the times. Image registration is based on the method of image feature points. This method is easy to use computer to handle and easy to realize human-machine interaction, and the focus is to how to extract image of characteristic points on effective. Elaborating image mosaicing technique for the purpose, significance, the domestic and foreign research present situation, development direction and the subject of the purpose and significance of the study, introducing the core technology of the image matching processing--Image registration.Expounding the existing registration method working principles, comparing and summaring advantages and disadvantages of all kinds of commonly used classical algorithms. Based on characteristics of the corner detection algorithm, introducing the Harris corner detection algorithm in detail. Use of MATLAB 7.0 software programming, realize image corner detection, image registration and image mosaicing, and elegant design graphical user interface (GUI), to demonstrate the final results. The simulation results prove that: Harris operator has a better adaptability, algorithm is simple and easy to program, realize image mosaicing with high stability and robustness,extract feature points well in the image with gray-scale changing, rotate existing noise and so on,and make corner detection by low mistake examining rate and high efficiency matc