Image modeling system development for robust descriptor of environmental change

Abstract

An image modeling system is critical tool for CDVS (Compact Descriptor for Visual Search) matching test in MPEG-7. There are uncountable environmental variables in real world and most application developers want to define as them. For such a reason, this paper first presents an image modeling system, and then applies the proposed system development for robust descriptor of environmental change. The system includes three core algorithms which are luminous source modeling development, illumination calibration development based on brightness distribution in reference images and geometrical modeling development. Experimental result shows that images are accurately processed as we want and gives the efficient test environment for matching.

6 Figures and Tables

Cite this paper

@article{Choi2013ImageMS, title={Image modeling system development for robust descriptor of environmental change}, author={Gveong Min Choi and Hvun Il Junz and Rae Kwang Kim and Weon Geun Oh}, journal={The 19th Korea-Japan Joint Workshop on Frontiers of Computer Vision}, year={2013}, pages={64-66} }