基于模糊灰度共生矩阵与隐马尔可夫模型的断口图像识别

纹理通常由空间分布和灰度分布共同描述，灰度共生矩阵(GLCM)能兼顾二者，故广泛应用于纹理分析中。在计算GLCM时，为降低其维数，需对纹理图像进行灰度量化，这必然丢失部分图像信息。灰度量化时，由灰度值与量化区间中心值的不同距离，构造出相应的模糊隶属度函数，并定义了模糊灰度共生矩阵(FGLCM)。通过对断口图像FGLCM的14个特征统计量进行相关性分析，选择角二阶矩和熵等7个统计量作为特征参数，并验证了其有效性。最后，在4类典型断口图像的特征空间上，采用隐马尔可夫模型(HMM)进行分类识别。实践表明，FGLCM比已有的GLCM能更好地表征断口特性,且在HMM状态数为3时，断口分类的平均识别率可达98%。

Recognition of fracture surface images based on fuzzy gray level co-occurrence matrix and hidden Markov model

Texture is usually depicted by a gray-level distribution along with a certain spatial interaction. Gray level co-occurrence matrix(GLCM) is an appropriate candidate to depicted texture because of its capability of blending spatial interaction with gray-level distribution, thus, it can be widely applied in texture analysis. When calculating GLCM, the gray-level quantization would be needed in order to decrease matrix dimension, and certain information would be lose. A membership function matrix is established whereby the distance which between the real gray-level and the mean of quantization gray-levels area, and then, a newly co-occurrence matrix, namely fuzzy gray level co-occurrence matrix(FGLCM) is proposed. After appropriate features are selected based on FGLCM statistics properties analysis, the hidden markov model(HMM) classification is applied to divide the classical fracture surface image to four kinds. It is proved practically that FGLCM in this paper is better than the GLCM in depicting textures and the FGLCM combined with HMM is efficient performance in fracture surface images classification, and the recognition rate is 98%.