Abstract: According to the demand for rapid construction of 3D battlefield environment, an automatic 3D battlefield environment construction technology was proposed based on UAV Remote Sensing Images. The specified area remote sensing images were captured using UAV The sparse reconstruction and calibrations were obtained using SFM method. The dense reconstruction was done using multi-view stereo matching algorithms. The scene surface was fit using dense points and texture mapping was done to generate 3D battlefield environment. The experimental results show that the proposed method can quickly reconstruct the battlefield, and improves the efficiency of 3D modeling. It has significant intelligence in terrain analysis, reconnaissance, combat assessment, etc.

Key words: UAV remote sensing image, structure and motion, dense point cloud, texture mapping, 3D battlefield