Share Email Print
cover

Proceedings Paper

The segmentation of texture surface under varying illuminant direction
Author(s): Jia-hui Cong; Yun-hui Yan
Format Member Price Non-Member Price
PDF $14.40 $18.00
cover GOOD NEWS! Your organization subscribes to the SPIE Digital Library. You may be able to download this paper for free. Check Access

Paper Abstract

When automatically inspecting textured surface defects, the most important step is to segment the defects from the background. Segmentation plays a crucial role in automated visual inspection. In this study, we present a method for segmentation of rough surface based on using texture feature derived from different illumination direction. Extract a set of feature images that described the characteristics of the textures. Variation in illumination direction effects image physical texture and then effects on texture segmentation. Previous studies have demonstrated that the surface topography as a cue for image segmentation. For complicated textures, however, two rough surfaces with different reflectance function, but similar topography must be distinguished. The images are often noisy and poor contrast. Compared with intensity, texture is more of a global property. A method is presented that to obtain images with maximal contrast by fusing a series of images which were acquired under different illumination directions. Using illumination series of images contain significantly more information about the surface characteristics than single images. The experiment result shows this is a robust and reproducible way to obtain high-contrast images containing the relevant information for subsequent processing steps.

Paper Details

Date Published: 15 April 2010
PDF: 6 pages
Proc. SPIE 7522, Fourth International Conference on Experimental Mechanics, 75224F (15 April 2010); doi: 10.1117/12.851659
Show Author Affiliations
Jia-hui Cong, Northeastern Univ. (China)
Yun-hui Yan, Northeastern Univ. (China)


Published in SPIE Proceedings Vol. 7522:
Fourth International Conference on Experimental Mechanics
Chenggen Quan; Kemao Qian; Anand Krishna Asundi; Fook Siong Chau, Editor(s)

© SPIE. Terms of Use
Back to Top