为实现快速和有效的织物瑕疵自动检测,提出了一种基于时间序列而不是图像的功率谱纹理分析方法。依据Burg au to regress ive(AR)算法估计得到谱数据,从中提取能够反映纹理周期和取向等特点的特征,并首次采用支持向量数据描述模型来检...为实现快速和有效的织物瑕疵自动检测,提出了一种基于时间序列而不是图像的功率谱纹理分析方法。依据Burg au to regress ive(AR)算法估计得到谱数据,从中提取能够反映纹理周期和取向等特点的特征,并首次采用支持向量数据描述模型来检测织物瑕疵纹理。对包含多种疵点的若干织物样本的检测结果表明,依照本文所述方案能够在保持较低的误警率前提下达到较高的疵点检出率,证明了所述方案的可行性。展开更多
Due to its advantages of objectiveness, automation, accuracy and fastness in various applications, the technology of computer vision has become one of the studying hotspots in the area of the objective inspection and ...Due to its advantages of objectiveness, automation, accuracy and fastness in various applications, the technology of computer vision has become one of the studying hotspots in the area of the objective inspection and assessment of textiles apparent properties during the past two decades in the world. Both a brief review of its applications in the recent decade both at home and abroad to the automatic inspection and assessment of the various apparent properties of the textiles, such as yarn, woven fabrics and knitting fabrics, carpet fabrics, nonwoven fabrics and textile webs, etc., and a detailed introduction to the research work including the objective evaluation of fabric wrinkle grade, automatic fabric defects detection and assessment of fabric pilling grade, etc., that was conducted by our research section, i.e., Computer Vision’s Textiles Application Research Section, College of Textiles, Dong Hua University, have been provided. Experimental results have proved the feasibilities of the approaches used by us in the applications to the objective inspection and assessment of fabric apparent properties, and also indicated that the technology of computer vision is a power tool for the objective and automatic inspection and assessment of textiles apparent properties, and that it has a bright application future.展开更多
文摘为实现快速和有效的织物瑕疵自动检测,提出了一种基于时间序列而不是图像的功率谱纹理分析方法。依据Burg au to regress ive(AR)算法估计得到谱数据,从中提取能够反映纹理周期和取向等特点的特征,并首次采用支持向量数据描述模型来检测织物瑕疵纹理。对包含多种疵点的若干织物样本的检测结果表明,依照本文所述方案能够在保持较低的误警率前提下达到较高的疵点检出率,证明了所述方案的可行性。
文摘Due to its advantages of objectiveness, automation, accuracy and fastness in various applications, the technology of computer vision has become one of the studying hotspots in the area of the objective inspection and assessment of textiles apparent properties during the past two decades in the world. Both a brief review of its applications in the recent decade both at home and abroad to the automatic inspection and assessment of the various apparent properties of the textiles, such as yarn, woven fabrics and knitting fabrics, carpet fabrics, nonwoven fabrics and textile webs, etc., and a detailed introduction to the research work including the objective evaluation of fabric wrinkle grade, automatic fabric defects detection and assessment of fabric pilling grade, etc., that was conducted by our research section, i.e., Computer Vision’s Textiles Application Research Section, College of Textiles, Dong Hua University, have been provided. Experimental results have proved the feasibilities of the approaches used by us in the applications to the objective inspection and assessment of fabric apparent properties, and also indicated that the technology of computer vision is a power tool for the objective and automatic inspection and assessment of textiles apparent properties, and that it has a bright application future.
