2009 Volume 49 Issue 5 Pages 709-718
An intelligent recognition and quantification system for photomicrographs of iron ore sinter is useful and convenient; however, it is impossible to develop a successful intelligent system without adequate and accurate texture features of mineralogical phases. The gray-level co-occurrence matrix (GLCM) has been proved as an effective method for extracting the texture features in other fields, therefore, this work examines texture features for the main mineralogical phases, such as magnetite and calcium ferrite, based on GLCM. These features include contrast, energy, entropy, and inverse difference moment. Specifically, this study addresses the effect on these features of several parameters, including the gray levels, the size of the image window, and the distance between the co-occurrences, and the offset angle. When the gray levels equal 125, the size of image window equals 100, and the distance of the co-occurrence equals 15, the average values of the four offset angles indicated that the features of each phase were relatively constant. Space distance characterizes the differences between a known image and an image to be analyzed; it determines the texture pattern of the image and is calculated using the Canberra space distance equation. Further calculation validates the results, indicating that intelligent recognition and quantification systems can be developed based on this method.