This paper describes three levels of fusing depth and intensity data, namely the pixel level, the primitive level and the matching level. Among the three, the primitive level fusion is the one closest to success in obtaining a single depiction by fusing the two data. Areas of further research and existing obstacles in enhancing the resolution of computer vision include manipulation and powerful reconfiguration; realization of parallel architecture; a comprehensive semantic labeling; object surface occlusion from a camera or laser projector.
|Journal||IEE Colloquium (Digest)|
|Publication status||Published - 1 Mar 1994|
|Event||Electronics Division Colloquium on 3D Imaging and Analysis of Depth/Range Images - London, UK|
Duration: 1 Mar 1994 → 1 Mar 1994