应用改进的灰度共生矩阵识别木材纹理多重特征值

目前木材的主要分类方式是由人的经验进行分类,分类的好坏取决于人的经验。因此机器自动检测分类木材种类变得迫在眉睫,目前机器识别木材种类最主要的方法是应用灰度共生矩阵(GLCM)提取木材纹理特征识别木材种类。但是基于灰度共生矩阵(GLCM)特征提取分类存在缺陷,这是由于木材图片旋转再识别时导致分类精度下降。本研究应用改进的灰度共生矩阵(I-GLCM)提取木材多重特征值,较前人提取的灰度共生矩阵(GLCM)识别木材种类,具有旋转不变性。应用matlab模式识别算法进行训练、分类。结果表明,应用本方法对木材进行分类,分类精度比应用灰度共生矩阵(GLCM)精度高,分类效果较好,是一种新的木材识别方法。

Application of Improved Gray Symbiosis Matrix to Identify the Multiple Characteristic Values of Wood Texture

There are many types of wood in China,and the correct classification of wood is essential.At present,the main classification of wood is primarily relied on human experience.Therefore,it is extremely urgent to use machine to automatically detect wood species.The most important method to achieve this is to use the gray level co-occurrence matrix (GLCM) to extract the wood texture features to identify the wood species.However,there are defects in GLCM,which is caused by the degradation of resolution when the wood picture is rotated and re- identified.Therefore,the modified gray level co-occurrence matrix (I-GLCM) is used to extract the multiple eigenvalues of wood,which is more invariant than the gray-scale co-occurrence matrix (GLCM) extracted by the predecessors.To solve the problem Matlab's pattern recognition algorithm was used for training and classification.The results showed that the classification accuracy was higher than that of applying GLCM to extract wood texture,and the classification effect was better.