一种图像校正重采样效应的评估及优化方法  

Evaluation and Optimization Method of Resampling Effects in Image Rectification

作　　者：宋汉辰[1] 吴玲达[1] 魏迎梅[1] 

机构地区：[1]国防科技大学五院指挥自动化系多媒体室,湖南长沙410073

出　　处：《小型微型计算机系统》2005年第8期1387-1390,共4页Journal of Chinese Computer Systems

摘　　要：分析了图像校正中重采样效应的主要表现形式,建立了一种重采样效应的评估方法.在此基础上,根据图像投影变换及校正图像上基础矩阵的性质,给出了两种类型的图像校正不变量,提出了一种独立于图像校正过程的重采样效应优化方法,通过投影区域修正的方法提高校正图像质量,实验证明,该方法能够有效的降低图像变形,减少图像分辨率的损失.Image rectification is the process of warping a pair of stereo images in order to align the epipolar lines with the scanlines of the images. Once a pair of images is rectified, stereo matching can be implemented in an efficient manner. However,many transformations will lead to unwanted image distortions. In this paper, we analysis the cause of effects of resampling and propose an evaluation method. Based on properties of fundamental matrix in image transformation and on rectified pairs, we give the definition of rectify invariables and propose two types of it to optimize the effects of resampling. The separating of rectification and correction widens the applicability bound. The method is proved to be efficiency in the experiments.

关 键 词：图像校正 重采样效应 校正不变量 基础矩阵 

分 类 号：TP317[自动化与计算机技术—计算机软件与理论]