The realistic of infrared texture has an important impact on the result of infrared scene simulation. Hence, the study of infrared texture is of great significance.Generation, real-time simulation application and confidence assessment of large-scale infrared ground background texture based on remote sensing image, in this paper, is studied. Firstly, ground background infrared texture is generated by combining physiognomy distribution from the visible remote sensing image with infrared radiation characteristic of physiognomy material. Secondly, infrared texture data is applied in the large-scale infrared scene simulation. Infrared image is real-time output by taking advantage of GPU programmable pipeline technique. Finally, SRMSE, Fidelity, Correlation and SSIM indexes are used to assess the similarity of infrared texture generated from visible image relative to actual captured infrared image. This method can provide the basic data source for infrared scene simulation, which has a wide range of applications. |