摘要
以系统的图像三维重构的立体匹配为主要研究内容,给出图像多尺度表示的数学模型.分析描述两幅图像之间约束的对极几何关系基本原理;研究了两幅图像稠密匹配的对极几何关系的代数表示形式,并建立了相应的数学模型;通过应用某两幅灰度数字图像进行仿真验算,证明了基于SIFT算法图像重构立体匹配方法的可行性,为相机系统识别目标在不同方位的特征信息提出了一种新的研究思路.
Based on system image 3 D reconstruction of stereo matching as the main research contents,the mathematical model of image multiscale representation is given. The basic principle of the constraint of epipolar geometry relationship between two images is analyzed. the algebraic representation of two images dense match is researched,and the corresponding mathematical model is established. By simulating and checking calculation of two gray images as examples,it is proved that the stereo matching of image reconstruction based on SIFT algorithm is feasible. It will provide a new research idea for camera system to recognize target＇s characteristic information in different directions.
出处
《西安文理学院学报（自然科学版）》
2017年第6期63-66,共4页
Journal of Xi’an University(Natural Science Edition)
关键词
三维重构
SIFT算法
对极几何关系
灰度图像
3 D reconstruction ： SIFT algorithm
epipolar geometry constraint
gray image
