Data fusion in web-based visualization of 3D scene semantic model

In the paper, the subsystem for graphical visualization of semantic messages has been proposed and configuration dataflow has been described. The key problem in 3D object rendering over video camera image is fusion of parameters collected during camera calibration and pose estimation with models of visualized objects. Pose estimation results and transformation matrix setup were described. Proposed system has been developed in HTML5 internet web page format that uses portable JavaScript language and WebGL library that make use of GPU device. Presented prototype can be used by a web browser on desktop PC as well as on mobile device. No third-party software is necessary.