Abstract—Estimation of the amount of fines in crushed rock from images using standard segmentation algorithms is difficult, owing to the heterogeneity and partial obscuring of the objects to be identified. In this paper, we present an image classification system to estimate the proportion of fines in aggregates of rock particles based on two-level wavelet decomposition and morphological operations. Morphological opening and closing filtering performed on rock images with structuring elements of increasing size also enhanced the class separability of the images of rock particulates. Experimental results showed that the performance of the image classification approach can be superior to standard methods.
Index Terms—Image-classification system, morphological operations, rock particulate size, wavelet.
Anthony Amankwah was with the Department of Process Engineering, University of Stellenbosch, South Africa. He is now with the School Computer Science, University of Witwatersrand, Private bag 3, Wits 2050, Johannesburg, South Africa (e-mail: Anthony.amankwah@wits.ac.za).
Chris Aldrich is with the Western Australian School of Mines, Curtin University of Technology, Perth, (e-mail: chris.aldrich@curtin.edu.au).
[PDF]
Cite:Anthony Amankwah and Chris Aldrich, "Estimation of Rock Particle Size Distribution Using Wavelet Decomposition and Morphological Operations," International Journal of Computer Theory and Engineering vol. 5, no. 3, pp. 450-453, 2013.
