DENOISING OF CORE PLUG IMAGE BASED ON NONLINEAR ANISOTROPIC DIFFUSION
UDC: 004.932:519.63
DOI: -
Authors:
Arseniyev-Obraztsov Sergey S.,
Zhukova Tatiana M.
Keywords: X-ray computer micro tomography, digital high-resolution microscopy, digital denoising of 3D images, nonlinear anisotropic diffusion filter, parallel algorithms, stencils on regular grids
Annotation:
A modified method of digital filtration based on nonlinear anisotropic diffusion is proposed. Filters using this approach have been successfully applied in medicine. This method was specifically modified to process the results of digital computer tomography and microscopy of core plugs, obtained from reservoirs of oil and gas fields. The filter allows to suppress both additive and multiplicative noise without changing the position of the internal boundaries of the object. Basing on the programming language supporting the “non-uniform memory access” (NUMA) paradigm the parallel filtration algorithm for large size 2D and 3D-digital images was developed for heterogeneous high-performance computer systems. The computer program operation is illustrated by the results of the analysis of core plug permeability dependence on the direction of the fluid flow.
Bibliography:
1. Арсеньев-Образцов С.С. Определение тензора коэффициентов проницаемости численным моделированием течения флюида на цифровой модели пористой среды//Труды РГУ нефти и газа имени Губкина. — 2015. — № 4. — C. 64-76.
2. Самарский А.А. Теория разностных схем. — М.: Наука, 1983. — 616 с.
3. De Boor C.: A Practical Guide to Splines. Springer Series: Applied Mathematical Sciences, vol. 27 1st ed. 1978. 1st hardcover printing, XVIII, 2001, 372 p.
4. Numrich R.W. and Reid J.K. Co-Array Fortran for parallel programming. ACM Fortran Forum. — 17(2). — 1998. — P. 1–31.
5. Perona P. and Malik J. Scale-space and edge detection using anisotropic diffusion. Proceedings of IEEE Computer Society Workshop on Computer Vision. — 1987. — P. 629-639.
6. Russ J.C. The Image Processing Handbook. CRC Press, Inc. — 2006. — 832 p.