MULTI-SENSOR FUSION FOR SIMULTANEOUS GEOMETRIC-PHYSICS MODELING OF ENVIRONMENT

This paper explores a multi-sensor system for simultaneous geometric-physical modeling of environment and its corresponding data processing methods. The system primarily utilizes millimeter-wave cascaded radar and multispectral camera to capture the electromagnetic feature and material categories of the target environment, while employing Light Detection and Ranging (LiDAR) and optical camera for three-dimensional (3D) spatial reconstruction of the target environment. To ensure the matching of the geometric-physical model and a more accurate modeling result, fusion of information obtained from the four sensors mentioned above is required. Finally, common environments are selected for the experimentation, and good spatial geometry-physics reconstruction effects are obtained.