Construction of large-scale 3D virtual environment is important in many fields such as robotic navigation, urban
planning, transportation, and remote sensing, etc. Laser scanning approach is the most common method used in
constructing 3D models. This paper proposes an automatic method to fuse image and laser-scanning data in a large-scale
3D virtual environment. The system comprises a laser-scanning device installed on a robot platform and the software for
data fusion and visualization. The algorithms of data fusion and scene integration are presented. Experiments were
performed for the reconstruction of outdoor scenes to test and demonstrate the functionality of the system. We also
discuss the efficacy of the system and technical problems involved in this proposed method.