基于IR-SFS算法空间目标红外影像3D重建

在利用单幅影像的明暗恢复形状（Shape From Shading，SFS）三维重构算法的基础上，提出同时考虑外界辐射源和目标自身红外辐射的IR-SFS（infrared-SFS）算法。首先通过分析SFS算法原理和空间目标的红外成像特性，建立IR-SFS辐射方程并进行了仿真研究，然后利用温度场估计获得红外差值图，在人工合成的理想半球体、半圆柱体卫星红外影像上进行算法测试，并以美国STS107真实红外影像作为实验目标进行三维重建。实验结果表明，所提出IR-SFS算法经过参数优化后，与原SFS算法相比，重建模型的峰值信噪比更高，对STS107顶部舱门、尾翼、机舱、机舱内方形部件具有更佳显示度，整体效果得到明显改善。

3D reconstruction of space target IR image based on IR-SFS algorithm

In this paper,a new 3D surface reconstruction algorithm named IR-SFS（ shape from shading for infrared images） for infrared images of space targets is proposed based on the traditional single image SFS（ shape from shading） algorithm,in which both external light source and self-radiation of infrared objects are considered. Firstly,the traditional SFS algorithm is described and the features of IR imaging of space targets are then analyzed to build the IR-SFS radiation equation followed by simulation study. Using temperature estimation to get the IR residual gray image,3D reconstruction models are achieved on the synthetic infrared images of he-mi-sphere,hemi-cylinder satellite and the real infrared image of USA STS107（ Space Transportation System Program 107） as the experimental targets. The experiment results show that higher peak signal to noise ratio（ PSNR） and better visibility at the top of the space shuttle door,the tail,the cabin and the square equipment in the cabin of STS107 can be obtained by using the proposed 3D reconstruction algorithm with optimized parameters. The overall performance is significantly improved compared to the original SFS algorithm.