点击次数：
所属单位：材料科学与技术学院
发表刊物：INTERNATIONAL JOURNAL OF ADVANCED MANUFACTURING TECHNOLOGY
关键字：X-ray NDT Image processing Noise-reduction algorithm Enhancement algorithm
摘要：Weldment X-ray image processing has great significance on the subsequent segmentation and extraction of weld seam and defects. In the current study, X-ray image for laser welding of aluminum alloy was adopted as experimental object. A new image denoising algorithm was proposed, which was combined with weighted adaptive median filter and noise-reduction algorithm based on wavelet transform. An improved adaptive fuzzy enhancement algorithm was put forward, which was based on traditional Pal-King fuzzy enhancement algorithm. Different noise-reduction algorithms were performed to denoise the X-ray image. Comprehensive evaluation of noise-reduction effect was conducted by comparing the processing effect, 3D grayscale distribution, and image quality evaluation index after different denoising algorithms. Moreover, fuzzy entropy and fuzzy index were used to estimate the enhancement effect of traditional Pal-King fuzzy enhancement algorithm and improved adaptive fuzzy enhancement algorithm. The results revealed that the noise-reduction effect and the image quality obtained by proposed algorithm were better than separately using weighted adaptive median filter or noise-reduction algorithm based on wavelet transform. Furthermore, after processing with improved adaptive fuzzy enhancement algorithm, the image detail information was more prominent, and sense of hierarchy was stronger.
ISSN号：0268-3768
是否译文：否
发表时间：2018-10-01
合写作者：Zhang, Dan,张丹,Yu, Haisong,Chen, Jie,陈杰,Li, Hao,魏艳红
通讯作者：魏艳红,占小红