Optimization method for performing texture mapping on non-Lambert surface based on consumer-level RGB-D sensor

The latest progress of sensor technology has introduced a low cost RGB video plus depth sensor, which can simultaneously acquire color and depth images at a video rate. Many researchers propose a method for performing texture mapping on a three-dimensional reconstructed model by using a consumer-level RGBD camera, such as Kinect. However, most existing texture mapping algorithms aim at Lambert models. The method is a texture optimization framework for the non-Lambert surface, depth information is fully utilized, and the surface texture of the model is reconstructed by using the synthesized image. The method has the following specific contributions: 1) an existing camera pose optimization strategy is expanded by utilizing depth information, and the accuracy in a non-texture region is improved; (2) color inconsistency between key frames is coordinated by using a joint color histogram, so that negative effects of highlight pixels on reconstructed textures are avoided; and 3) providing a BDS-G texture synthesis m