Summary
|
For Full-Text PDF, please login, if you are a member of IEICE, or go to Pay Per View on menu list, if you are a nonmember of IEICE. |
|
Color Texture Segmentation Using Color Transform and Feature Distributions Shiuh-Ku WENG Chung-Ming KUO Wei-Cung KANG Publication IEICE TRANSACTIONS on Information and Systems Vol.E90-D No.4 pp.787-790 Publication Date: 2007/04/01 Online ISSN: 1745-1361 DOI: 10.1093/ietisy/e90-d.4.787 Print ISSN: 0916-8532 Type of Manuscript: LETTER Category: Pattern Recognition Keyword: color index histogram (CIH), local binary pattern (LBP), boundary checking, Full Text: PDF(936.7KB)>>
Summary: This letter presents a simple scheme to transform colors to some representative classes for color information reduction. Then, the weighted distributions of color index histogram (CIH) and local binary pattern (LBP) are applied to measure the similarity of adjacent texture regions during the segmentation process. In addition, for improving the segmentation accuracy, an efficient boundary checking algorithm is proposed. The proposed method not only saves execution time but also segments the distinct texture regions correctly. | ||||||||||
| ||||||||||
