JOGCE Navigation
Publication Ethics
Call For Paper
Guide to Authors
Editorial Board
Submit Manuscript
Viewing Options
[View Abstract]
[View Full Text PDF]
[View Full Text PDF]
Authors Articles on Google Scholar
Sumanta Das*
Malini R Choudhury
Authors Articles on Pubmed
Sumanta Das*
Malini R Choudhury
Email this Article to a friend
Print this Article
Viewed: 902
Forwarded: 0
Printed: 169
Downloaded: 1,966
Browse Journals By Category
Agriculture and Food Sciences
Biological Sciences
Education and Arts
Environmental Sciences
Medical Sciences
Pharmaceutical Sciences
Physical and Natural Sciences
Social Sciences

Newsletters Subscription

Journal of Oil, Gas and Coal Engineering

Rock type classification by image analysis using the quaternion colour extraction model and support vector machine classifier

Sumanta Das and Malini Roy Choudhury

Accepted 14 April, 2014.

Citation: Das S, Choudhury MR (2014). Rock type classification by image analysis using the quaternion colour extraction model and support vector machine classifier. J. Oil Gas Coal Engin. 1(1): 002-009.

Copyright: © 2014 Das and Choudhury. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are cited.


In mineral processing plants it is important to estimate rock composition, size and grindability to improve control of the grinding process. This paper proposes a new method and extends a general remote sensing approach for estimation of rock type estimation using colour information. Our proposed method requires the division of each image into sub-images using the Binary Quaternion-Moment-Preserving (BQMP) for colour feature extraction and support vector machines (SVM) for classification. The colour feature extraction BQMP method splits the image in two and chooses representatives of each half using the histogram as features. Once the feature vector has been computed, each vector is assigned to one of the classes using a classifier. The method was tested on two databases. The first one, contains images from rocks obtained from a nickel mine with three classes and the second database contains images from rocks obtained from a copper mine in India with seven rock types. The classification accuracy was compared based on a mixture of wavelet texture analysis (WTA) for texture feature extraction and principal components analysis (PCA) for colour features. WTA-PCA approach reached 78.8% accuracy in the test set of the first database and 69.3% with the second database, while BQMP method reached 93.05% and 91.8% of classification accuracy respectively.

Keywords: Rock and lithological classification, grindability estimation, BQMP features extraction, SVM, colour features.