期刊
AUTOMATION IN CONSTRUCTION
卷 31, 期 -, 页码 338-356出版社
ELSEVIER SCIENCE BV
DOI: 10.1016/j.autcon.2012.11.003
关键词
Coating defect recognition; Color image processing; Fourier transform; Texture feature
Most acquired steel bridge coating images are classified into two groups, defect and non-defect, by naked eyes or using the method reported in Lee's work [18]. Lee's method works effectively with blue-paint-coated steel bridge images in the State of Indiana, U.S.A. However, its effect on other coating colors or in some particular conditions, such as coating images with background noise or non-uniform illumination, has not yet been explored. In view of this, a rust defect recognition method based on color and texture feature (RUDERM), which combines the Fourier transform and color image processing, is proposed in this research. RUDERM aims to adapt to various background colors and overcome the influences from particular conditions. After comparisons of processed results, it is proven that RUDERM has an advantage in handling non-uniform illumination and can achieve a shorter processing time, which may lead to real-time coating inspection in the near future. (C) 2012 Elsevier B.V. All rights reserved.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据