采用部分灰度压缩扩阶共生矩阵的煤和煤矸石图像识别

提出一种基于部分灰度压缩扩阶共生矩阵的煤和煤矸石图像识别方法.首先,对煤和煤矸石0~255级灰度图像的前部分灰度信息作灰度级压缩和灰度矩阵扩阶处理,对剩余灰度级部分保持原灰度级不变;然后,根据灰度共生矩阵纹理特征分析理论,分别计算压缩扩阶后的煤和煤矸石灰度图像的能量、熵、惯性矩及相关性.最后,对煤和煤矸石各100张样本采集图像进行处理,并依据特征参数分类识别.结果表明:基于部分灰度压缩扩阶共生矩阵的特征参数能够很好地对煤和煤矸石图像进行有效识别,总的正确率达到93.5%.

Image Recognition Method of Coal and Coal Gangue Based on Partial Grayscale Compression Extended Coexistence Matrix

A coal and coal gangue image recognition method based on partial grayscale compression extended coexistence matrix is presented. Firstly, the 0-255 grayscale images of coal and coal gangue are compressed partly with the front part grayscale, while the other parts of grayscale are remained the same with the original grayscale. Then, according to texture analysis theory of gray-level co-occurrence matrix(GLCM), the energy, entropy, moment of inertia and the correlation coefficient of the coal and coal gangue after compression and extension-order are calculated,respectively. The experiments are carried out with the test samples of 100 coal images and 100 coal gangue images, and the performances of the proposed recognition method are demonstrated with the calculated characteristic parameters. The experimental results indicated that the coal and coal gangue images can be recognized effectively, and an overall accuracy up to 93.5% is achievable with the proposed expanded-order GLCM method.