织物图像增强矩阵特征模型的建立

Establishment of enhanced matrix model for fabric image feature

针对目前缺乏有效显现织物特征的成熟模型,使织物疵点识别效果不佳的现状,提出一种新的织物图像特征模型,即增强矩阵特征模型。该特征模型以图像的灰度值为基础,引入一种新的增强矩阵。该矩阵由根据织物图像梯度变化生成的矩阵算子组成,可对像素灰度值进行变换计算以放大或缩小图像局部特征,使图像的特征显现更加层次分明。通过采用MatLab 7.0编写程序验证该特征模型,发现该模型对织物疵点特征的显现效果明显,使疵点区域的特征变异得到明显增强,为织物疵点识别提供了一种新的思路。

Since the workable model for effective identification of fabric features are not available yet,the result of fabric defect recognition needs to be improved.Thus,this paper has proposed a new fabric image feature model,i.e.,＂enhanced matrix feature model＂,which introduces a new enhanced matrix based on the gray of the image.The matrix is made of matrix operators generated by gradient variation of fabric image,which can do transform calculation for pixel value to amplify or reduce the local characteristics of the image so that the image features are displayed clearly.Some programs are compiled to verify this model using MatLab 7.0,and the verification result shows this model is suitable for defect recognition of fabric feature,because the feature variations in the defeat area have been enhanced remarkably.This model has provided a new idea for fabric defect recognition.