基于生物信息学中双DNA序列比对算法的图像立体匹配及其实现  

Novel stereo matching algorithm based on pair-wise DNA alignment algorithm in bioinformatics and its implementation

作　　者：谢少荣[1] 王东红[2] 罗均[1] 龚振邦[1] 

机构地区：[1]上海大学机电工程与自动化学院,上海200072 [2]广西财经学院,广西南宁530002

出　　处：《光学精密工程》2007年第1期106-111,共6页Optics and Precision Engineering

基　　金：国家自然科学基金资助项目(No.50405046和No.60605028);上海市科委资助项目(No.045107031);上海市优秀青年教师培养计划资助项目(No.04Y0HB094);上海大学优秀青年教师后备人选科研项目

摘　　要：提出了一种基于生物信息学中双DNA序列比对算法的图像立体匹配新方法。图像立体匹配和生物信息学中双DNA序列比对的实质都是在匹配准则下搜索最佳匹配基元,因而新颖地将双序列比对算法引入图像立体匹配。首先介绍了基于动态规划的双序列比对算法原理及其用于图像立体匹配的实现方法,然后根据左右摄像机的最大视差是一个有限定值,进行了算法改进,极大地减少了计算量,并给出了VC6.0中的实现流程,最后采用4组不同的图像对进行了实验验证。该方法具有较低的计算复杂度和适宜于并行计算的特点,生成的视差图效果表明双序列比对算法为图像立体匹配提供了一个实用有效的方法。A novel stereo matching algorithm based on pair-wise DNA alignment algorithm is presented. The essential of both stereo matching and pair-wise DNA alignment in bioinformatics is that the correspondence points are searched by matching criteria, so the pair-wise DNA alignment algorithm is introduced to design a new stereo matching algorithm. Firstly, the principle of the dynamic programming and implementation of the proposed algorithm are presented. Then, this algorithm is significantly improved to reduce the calculation drastically, because there is a maximum possible disparity whose value can be derived from the field of view of the cameras, .the physical distance between the two cameras, and the focal length of the cameras. The flow of the algorithm is designed in detail with VC6.0. Finally, the disparity maps of several different test images by means of this algorithm are shown,the advantages are low computer complexity and parallel processing. The results show that the proposed algorithm is useful and effective.

关 键 词：立体匹配 立体视觉 DNA序列 双序列比对 对应点 

分 类 号：Q-334[生物学] TP391[自动化与计算机技术—计算机应用技术]