基于遗传算法的遥感图像配准

【作者】 梅小明；

【导师】 陈绍裘；

【作者基本信息】 中南大学 ， 地球探测与信息技术， 2004， 硕士

【摘要】 图像配准指的是把一个图像区域从另一个往往由别的传感器摄取的相应图像区域中确定出来或找到它们之间对应关系的一种重要的图像处理与分析技术。图像配准技术广泛应用在导航定位、目标识别、运动分析、立体视觉、数据融合、变迁检测中，是这些图像分析技术的基础。图像配准问题的困难之处，在于图像类型的复杂多变和两幅图像之间的各种成像畸变。寻找适应性强、精度高、计算快的配准算法一直是研究的热点问题。而遥感图像配准指的是将取自同一目标区域的两幅或多幅图像在空间位置上最佳匹配起来，这些图像或者是由同一传感器在不同时相获取的，或者是来自不同的传感器，属于图像配准的范畴。和图像配准性质相同，遥感图像配准问题的核心也是需要找到两幅图像之间的最优对应关系或数学变换，这是一个典型的参数寻优问题，能够解决这一问题的算法有很多种，遗传算法是众多算法中的一种。本文从基于遗传算法的遥感图像配准方面展开研究，探讨、研究遗传算法用于图像配准的可能性，围绕这一目标，本文分成以下几部分进行阐述： 论文的第一部分，详述图像配准技术的基础理论。首先，论述了图像配准问题的数学描述和本质；其次，分析了影响配准性能的因素，提出了成像畸变在图像灰度分布上的差异可分为四类的观点；然后，给出了配准算法的四个构成因素。并对图像配准算法进行了综述，形成了图像配准技术的理论基础，也是本文的基础。 论文的第二部分，简单介绍了遗传算法，阐述遗传算法的运算过程及构成因素，并对其特点进行了分析。是本文算法的基础。 论文的第三部分，针对遥感图像的配准这一中心目标，用有别于传统的图像配准算法的遗传算法对此进行了试验。并详述了遥感图像文件的读写和遗传算法的优化策略。对遗传算法和相关系数法进行了性能比较。 在本文的结束语中，作者给出了对进一步研究工作的一些个人看法和建议。更多还原

【Abstract】 Image registration is an important analysis and processing technique, which located a image area from another image area or find out the correspondence between them, where the images taken at different times, by different sensors, or by different viewpoints from the same scene. Image registration technique is used in a wide-range areas, such as navigation and location, object recognition ,movement analysis, stereo vision, data fusion, change detection. It is the base of these images processing techniques. The difficulty of designing registration algorithm lies in the variety of scene styles and variety of imaging variations between images. How to design an algorithm which works with high adaptability, high precision and rapid computing efficiency is the kernel study topic. And remote image registration which belonging to image registration is a technique which best matching two or more images at space place from the same area, where the images taken at different times or taken by different sensors. Similitude to image registration, the key problem of remote image registration is how to find out the best correspondence or mathematic transformation between two images, this problem is a kind of typical problem which finding the best parameters, and genetic algorithm is one of the algorithms which can solve the problem. In this paper, the study of remote image registration based on genetic algorithm is had. According to the goal which discussing and studying the probability, the study is spread in the three parts of this paper.In the first part of the paper, image registration problem as a whole is analysed and discussed. The mathematics description of image registration problem, factors the affect registration performance, framework to design image registration algorithm including feature space, similarity merits, search space and search algorithm are detailed sequentially. The ideas to divide the variation into four styles and divide a registration algorithm into four factors are the director to design a matching algorithm. The theory of image registration is the basis of this paper.In the second part of this paper, genetic algorithm is briefly introduced, the processing and the factor of it is also expatiated, And the trait of it is discussed. Genetic algorithm is the basis of the algorithm used in the paper.In the third part of the paper, under the goal of remote image registration, genetic algorithm which differing from traditional algorithms is tested, following this, the reading and writing of remote image is detailed. According to the specific problem, genetic algorithm is designed and optimized. Finally, genetic algorithm and correlation method are compared. This part is the emphases and difficulty part of the paper.Finally, the author’s ideas and advices to extend study are given in the chapter of tag.更多还原