Rock Image Segmentation Based on Wavelet Transform and Watershed Algorithm
In order to obtain information of the reservoir micropore, and extract its spatial topological structure, wavelet transform and watershed algorithm are combined to segment rock images. Firstly, wavelet hard threshold algorithm and wavelet soft threshold algorithm are applied, respectively, to remove rock image noises. Secondly, watershed algorithm is used to segment rock images that not being denoised, denoised by wavelet hard threshold algorithm, and wavelet soft threshold algorithm for characteristics extraction of microscopic pore. Lastly, application results of these three methods for segmentation of rock images are compared and analyzed. The comparative results show that the best segmentation method for rock images is the combination of wavelet soft threshold algorithm and watershed algorithm, the combination of wavelet hard threshold algorithm and watershed algorithm is better, and the single watershed algorithm is the worst. The comparative results also show combination of wavelet soft threshold algorithm and watershed algorithm could segment rock images and extract their spatial topological structures accurately. The rock image segmentation based on wavelet soft threshold algorithm and watershed algorithm overcomes shortcomings of other methods and lays an important foundation for 3D reconstruction of the rock pore structures.
KeywordsRock image Denoising Segmentation Wavelet transform Watershed algorithm
Supported by Shaanxi National Science Foundation (No. 2012JQ8040) and National Training Program of Innovation and Entrepreneurship for Undergraduates (No. 201510705233).
- 1.Wen CJ, Wang SS, Yu HL et al (2013) Image segmentation method for maize diseases based on pulse coupled neural network with modified artificial bee algorithm. Trans Chin Soc Agric Eng 29(13):142–149Google Scholar
- 3.Wang T, Ji ZX, Sun QS (2015) A segmentation algorithm combined with non-local information and graph cut. J Comput Aided Des Comput Graph 27(5):783–791Google Scholar
- 4.Liu JP, Chen Q, Zhang J et al (2016) Interactive image segmentation based on ensemble learning. Acta Electron Sin 44(7):1649–1655Google Scholar
- 5.Xu QY, Li Y, Lin WJ et al (2017) Remote sensing image segmentation based on information clustering. J China Univ Min Technol 46(1):209–214Google Scholar
- 6.Ye RQ, Niu RQ, Zhang LP (2011) Mineral features extraction and analysis based on multiresolution segmentation of petrographic images. J Jilin Univ Earth Sci Ed 41(4):1253–1261Google Scholar
- 7.Yang YM, Liu N, Cheng GJ et al (2016) Clustering analysis of rock images based on Spark platform. J Xi’an Shiyou Univ Nat Sci Ed 31(6):114–118Google Scholar
- 9.Xu Y, Weaver J, Healy M (1999) Wavelet transform domain filters: a spatially selective noise filtration technique. IEEE Trans Image Process 3(6):747–758Google Scholar
- 11.Li M, Feng XC, Yang WJ (2006) Image denoising based on total variation and hard wavelet shrinkage. Signal Process 22(6):917–919Google Scholar
- 12.Lin ZJ, Yan ZJ, Xiao M et al (2015) Signal denoising based on empirical mode decomposition and wavelet soft threshold. J Qingdao Univ Sci Technol Nat Sci Ed 36(4):463–467Google Scholar