面向织物疵点检测的图像处理技术研究进展

随着对纺织工业产品质量要求的提高以及传统疵点检测方法存在局限性,基于图像处理技术的织物疵点自动检测技术得到了快速的发展。为提高图像处理技术的应用效率,实现纺织行业的数字化与智能制造,介绍了织物图像的预处理技术,对织物疵点检测的主流方法进行了总结,包括基于结构、统计、频谱、模型和学习的方法,并对这些方法的检测原理做了概括,分析了其优缺点与适用范围;介绍了现有成品检测设备,对比分析了仪器和系统处理技术的优缺点;最后,梳理分析了现有的图像处理技术在纺织工业应用中所面临的难题,并提出了对未来发展的构想。

Research progress of image processing technology for fabric defect detection

With the enhancement of product quality requirements in the textile industry and the limitations of traditional defect detection methods, the automatic detection of fabric defects based on image processing technology has seen an rapidly development. Compared with traditional technology, the application of image processing technology improves the processing efficiency and realizes the digitization and intelligent manufacturing of the textile industry. This paper introduces the preprocessing technology of fabric images, and summarizes the mainstream methods of fabric defect detection, including structure-based methods, statistics-based methods, spectrum-based methods, model-based methods and learning-based methods. In addition, it reviews the principles of these methods, and examines their advantages and disadvantages and scope of applications. Besides, the paper introduces the existing finished equipment and compares the advantages and disadvantages of these equipment. Difficulties facing the existing image processing technology in the application of the textile industry are analyzed, and prospects of future development are discussed.