论文部分内容阅读
大视场景象是大视场摄象机摄得的图象，它存在着严重的非线性几何失真（桶形失真）。为了提高匹配算法的定位精度和匹配概率，对摄得的实测图象进行几何失真校正的预处理是十分必要的。本文讨论了几何失真图象的象素位置与镜头焦距的关系，采用虚拟数字成象的方法，实现了几何失真校正。文中给出了变焦距地址修正的具体算法（ＺｏｏｍＡｄｄｒｅｓｓＭｏｄｉｆｉｃａｔｉｏｎａｌｇｏｒｉｔｈｍＺＡＭＡ），并对一幅实测图象进行了几何校正实验，取得了理想的校正结果。
Large field of view is a large field camera images taken, it exists a serious nonlinear geometric distortion (barrel distortion). In order to improve the positioning accuracy and matching probability of the matching algorithm, it is necessary to preprocess the measured distortion of the measured image. This paper discusses the relationship between the pixel position and the lens focal length of the geometrical distorted image, and realizes the geometric distortion correction by using the virtual digital imaging method. In this paper, ZoomAddressModificationalgorithmZAMA is given, and a geometric correction experiment is carried out on a measured image, and an ideal calibration result is obtained.