Application of Image Texture Retrieval in Mineral Discrimination
改进的图像纹理检索方法在矿石识别中的应用

The paper depicts mineral discrimination through image texture retrieval.The whole procedure goes like this: firstly dividing a large image into small window images,whose feature values are then extracted according to cooccurrence matrices and gray level run length matrices,with the limit of practical computation capability,only one single feature vector is applied in the classification.As we can see in the paper,a satisfying consequence can be reached when an appropriate threshold value can be provided.