Comprehensive assessment of non-uniform illumination for 3D heightmap reconstruction in outdoor environments

Three-dimensional (3D) surface textures can display texture information of an object better than their 2D counterparts, as 3D appearances can be varied with the scene reflected illuminations and view angles. Photometric Stereo, as one of the most effective technologies for capturing 3D surface information, has attracted wide attention. Uniform illumination is an essential condition for reconstructing surface heightmaps. In practical applications, non-uniform illumination usually leads to distorted heightmaps when reconstructing surfaces in outdoor environments. To tackle this problem, this paper presents three useful approaches including an improved self-adaptive method, a FSAM (fast self-adaptive method) and a manual correction method, to correct non-uniform illumination and eliminate the distortions for the reconstruction of 3D surface heightmaps. Experimental results show that these three methods are promising and effective in real-world outdoor environments.