期刊名称：International Journal of Multimedia and Ubiquitous Engineering
印刷版ISSN：1975-0080
出版年度：2014
卷号：9
期号：1
页码：403-414
出版社：SERSC
摘要：In this paper, we propose an intelligent method to extract flaws from nondestructive testing(NDT) images of ceramics. Our goal is to extend the previous study[8]'s feasibility to handle images having skewed brightness distribution or darker images. Thus, we use Gaussian filtering along with anisotropic filtering and sigma fuzzy binarization instead of repetitive binarization used in [8]. These techniques enable us to enhance the brightness contrast successfully so that the usability of our method is extended and verified in experiment. Two phases of image processing - area segmentation and extraction of the defect by searching for labeled pixel to form a defected object by Grassfire algorithm - are explained in detail
