织物疵点识别算法的研究

【作者】 于坤；

【导师】 韩其睿；

【作者基本信息】 天津工业大学 ， 计算机应用技术， 2008， 硕士

【摘要】 在纺织工业中,质量的控制是非常重要的,而织物疵点的检测是其中最重要的一部分。疵点检测的目的是在织造和验布过程中及时发现疵点,通过修复和整理,尽可能降低由疵点导致的织物质量的下降。本课题在对国内外织物疵点自动检测系统和理论成果学习与研究的基础上,主要利用数字图像处理与分析技术对缺经和缺纬两类织物疵点进行了研究,改进了识别算法,提高了识别速度,减少了计算时间。同时,采用VC++与Matlab混合编程的模式,提高了软件开发的效率和软件运行速度。本课题所做的主要工作为:将利用硬件设备得到的原始织物图像输入到计算机,然后将原始织物图像转换为灰度图像,再进一步转换为索引图像。利用3×3窗口的中值滤波器对图像进行去除噪声的处理,再利用直方图规定化增强图像对比度。在完成图像预处理的基础上,对图像进行快速检测,判断该图像是否含有疵点,对于含有疵点的图像进行小波变换,得到垂直和水平细节图像,并进行归一化得到织物各个特征值的曲线。将带疵点的织物图像的特征值曲线和正常织物的特征值曲线进行比较得到不同的特征值对不同类型疵点的响应是不同的,从而判断出疵点的存在。通过对运行结果分析,本课题提出的疵点识别算法能够准确检测平纹机织物中的缺经、缺纬的疵点。更多还原

【Abstract】 In the textile industry, quality control is very important, and the examination of fabric is an important part. The purpose of the fabric defects identification is to discover them in the time in the weaving and examing of the cloth process and reduce the fabric quality drop which causes by the defects as far as possible through the repair and the reorganization.Based on the study and research of the development of the automatic deceting technology, the detection of fabric defects by digital image processing and analyzing technique is studied,two kinds of fabric defects end out and miss pick are researched in the thsis, then improve the identification algorithm,enhance the identification speed,reduce the computing time. At the same time,to make the program run fast and more effective in the developing,the co-programming with VC++ and Matlab is used in this thesis.The main work done in the thsis : primitive fabric images obtained by the hardware equipment are input to the computer,then the primitive fabric images will be converted to grayscale images, further converted to index image, and then use the 3×3 window in the value of image filter to remove noise treatment, reuse of the histogram provides enhanced image contrast. Completed on the basis of image preprocessing, the image rapid detection, the judgement of the presence of faults image, the image containing defects wavelet transform, the details are vertical and horizontal images, and normalization of fabric are all eigenvalues curve , and the fabric belt defect image and the characteristics of the normal curve of the fabric eigenvalue curve comparisons of different characteristics of different types of faults on the response is different, and thus judged the existence of a defect.Through the analysis of the results, the defects identification algorithm proposed in this thesis can exactly detect the defects of end out and miss pick on the tabby fabric.更多还原