Texture analysis and classification for clinker in rotary kiln

This paper presents a potential application of texture analysis and classifier for a set of sintered clinkers in rotary kiln. This approach, which consists of the extraction of clinker from the whole image, GLCM texture features computation, and C4.5 classification, provides an efficient way for untouched on-line analysis of sintered clinker for the automatic control of industrial rotary kiln. 69 clinker images from an alumina rotary kiln were employed, 5 sets of 14 texture features were computed based on GLCM of different distances, texture feature classification were achieved by C4.5 algorithm, all graphic illustrations of classification rule are given as a function of texture ID. Experimental results demonstrate that it is effective to classify these sintered clinkers of different qualities. This paper may provide a basis for analyzing other kinds of industrial materials in rotary kiln.