航空图像的三维目标图像重建过程的仿真分析

Aviation Image Simulation Analysis of 3D Object Reconstruction Process

随着被飞行器摄体的越来越远,航空图像的拍摄距离往往较远,使得采集得到的图像特征变得扁平。形成反向透视畸变。传统的航空图像的三维目标图像重建方法针对这种透视畸变的干扰,多采用增加航拍图像进而增加图像特征的方法抵消干扰误差,但是,随着航拍图像特征的增加,也增加了重建过程中,特征校验与重建的负担,缺陷较为明显。提出了仿射近似算法的三维图像重建方法,利用仿射投影模型近似透视投影模型,运用线性算法获得透视畸变未知数的初始值,在使用透视投影模型对得到的三维重建数据深入优化来提升精度,提高了重建速度与精度。仿真结果表明,利用仿射近似算法进行航空图像的三维目标重建,提高了精度与效率。

A 3D image reconstruction method based on affine approximation algorithm is proposed. The affine projection model is used to approximate the perspective projection model,and the linear algorithm is used to obtain the initial value of perspective distortion unknown number. Then,the perspective projection model is applied to make deep optimization of obtained 3d reconstruction data to improve the accuracy,reconstruction speed and precision.Simulation results show that the proposed affine approximation algorithm for 3D object reconstruction of aviation image can improve the accuracy and efficiency.