基于Cult3D的倾斜摄影图像前景自适应配准方法  

Adaptive Foreground Registration Method for Tilt Images Based on Cult3D

作　　者：陈进强[1] 翁存福 CHEN Jin-qiang;WENG Cun-fu(Gongqing College of Nanchang University,Jiujiang Jiangxi 332020,China;Scienceand Technology College,Jiangxi Normal University,Jiujiang Jiangxi 332020,China)

机构地区：[1]南昌大学共青学院,江西九江332020 [2]江西师范大学科学技术学院理,江西九江332020

出　　处：《计算机仿真》2022年第11期267-270,298,共5页Computer Simulation

摘　　要：受到倾斜摄影前景图像重复纹理的影响,现有配准方法在运用过程中,产生一定的噪声灰度值,导致配准效果受限。针对上述问题,研究一种基于Cult3D的倾斜摄影图像前景自适应配准方法。使用Moravec算子将倾斜摄影前景图像处理为一个方形窗口,引入高斯函数构建图像前景倾斜转换对应关系,根据灰度数值大小,以Cult3D搭建配准的基元控制前景图像中的重复纹理特征,将子门限数值整合排列为一个数值路径,不断逼近匹配待配准的配准点,最终实现配准方法的自适应优化。设定倾斜图像的变换参数,进行仿真的结果表明,噪声灰度值得到有效抑制,图像配准过程的精度、速度较大,适合实际处理倾斜摄影图像。Affected by the repeated texture of the foreground image of oblique photography,the existing registration methods produce a certain noise gray value in the application process,resulting in limited registration effect.Aiming at the above problems,a method of adaptive foreground registration of oblique photographic images based on Cult3 D was studied.Moravec operator was introduced to convert the foreground image of oblique photography into a square window.Gaussian function was used to construct the corresponding relation of image foreground tilt conversion.Based on the gray value and Cult3 D to build the registration primitives,the repetitive texture features in the foreground image were controlled.The sub threshold values were integrated and arranged into a numerical path,infinitely approaching the registration points to be registered.Finally,the adaptive optimization of the registration method was completed.The simulation results show that the noise gray value is effectively suppressed,and the accuracy and speed of image registration process are greatly improved,indicating that this method is suitable for processing oblique photographic images.

关 键 词：倾斜摄影 图像前景 自适应配准 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]