Title :
A new specular highlights detection algorithm using multiple views
Author :
Ching, Wee-Soon ; Toh, Peng-Seng ; Er, Meng-Hwa
Author_Institution :
Sch. of Electr. & Electron. Eng., Nanyang Technol. Univ., Singapore
Abstract :
Specular highlights detection in a fundamental problem in the area of image processing and computer vision. In its formulation, the authors´ algorithm combines two important properties of specular highlights, namely, the strong peak in the image intensity histogram and the reduction in the best-matched correlation coefficient when matching images taken from multiple view points. Experimental results show that the proposed algorithm greatly enhances the reliability of detecting specular highlights over the conventional gray-scale specular highlights detection algorithms
Keywords :
computer vision; optical correlation; algorithm; computer vision; correlation coefficient; image intensity histogram; image processing; multiple view points; reliability; specular highlights; Computer vision; Detection algorithms; Erbium; Focusing; Gray-scale; Histograms; Image processing; Layout; Optical reflection; Pigments;
Conference_Titel :
Networks, 1993. International Conference on Information Engineering '93. 'Communications and Networks for the Year 2000', Proceedings of IEEE Singapore International Conference on
Print_ISBN :
0-7803-1445-X
DOI :
10.1109/SICON.1993.515809