2518-1092Research result. Information technologies2518-109210.18413/2518-1092-2016-1-3-4-9777COMPUTER SIMULATIONHIGH-PERFORMANCE ANALYSIS METHOD AND MORPHOLOGICAL IMAGE PROCESSINGHIGH-PERFORMANCE ANALYSIS METHOD AND MORPHOLOGICAL IMAGE PROCESSINGSoynikovaEkaterina SergeevnaSoynikovaEkaterina Sergeevna831468@bsu.edu.ruRyabykhMaxim SergeevichRyabykhMaxim Sergeevich828130@bsu.edu.ruBatishchevDenis S.BatishchevDenis S.batishchev@bsu.edu.ruSinyukVasily GrigorievichSinyukVasily Grigorievichvgsinuk@mail.ruMikhelevVladimir MikhailovichMikhelevVladimir Mikhailovichmikhelev@bsu.edu.ru20161300

Segmentation is a difficult stage in the processing and analysis of medical images. This is due to the high variability of their characteristics, low contrast processed images and the organization of complex geometric objects. The article covers the realization of Sobel operator and Canny algorithm using OpenMP parallel programming technology and NVIDIA CUDA. It is shown that the implementation of these algorithms for GPUs with CUDA technology improves imaging performance. Completion of the computational experiment showed the effectiveness of the implementation of Canny algorithm using CUDA technology, compared with OpenMP for different resolutions of medical images.

Segmentation is a difficult stage in the processing and analysis of medical images. This is due to the high variability of their characteristics, low contrast processed images and the organization of complex geometric objects. The article covers the realization of Sobel operator and Canny algorithm using OpenMP parallel programming technology and NVIDIA CUDA. It is shown that the implementation of these algorithms for GPUs with CUDA technology improves imaging performance. Completion of the computational experiment showed the effectiveness of the implementation of Canny algorithm using CUDA technology, compared with OpenMP for different resolutions of medical images.

segmentationmedical imagingselection bordersSobel operatorCanny algorithmGPUOpenMPNVIDIA CUDAsegmentationmedical imagingselection bordersSobel operatorCanny algorithmGPUOpenMPNVIDIA CUDAСписок литературыR. Gonzalez, R. Woods, Digital Image Processing [text] / R. Gonzalez, R. Woods; M.: Technosphere, 2005. 1072 p.Batishchev D.S, Mikhelev V.M The Infrastructure of High-performance Computer System for the Implementation of Cloud Storage and Analysis of Personal Medical Data. Scientific Bulletin of Belgorod State University. Series: Economy. Computer Science. 2016. Vol. 37. № 2 (223). Pp. 88-92.Ryabykh M.S. Soynikova E.S. Batishchev D.S., Mikhelev V.M. Meeting the Challenge of Segmentation of Medical Images Using Computing on Graphics Processors. Prospects of Development of Information Technologies. 2016. № 29. Pp. 163-168.Forsyth D.A., Computer Vision. The Modern Approach [Text]. / D.A. Forsyth, J. Ponce; transl. from English. M.: Publishing House "Williams", 2004. 1006 p.Y. Duan, J. Wang, M. B. T. Kam, and J. F. Canny, "Privacy preserving link analysis on dynamic weighted graph," Computational & Mathematical Organization Theory, vol. 11, no. 2, pp. 141-159, July 2005.J. F. Canny, "A computational approach to edge detection," IEEE Trans- ac ons on Pa ern Analys s and Mach ne In ell gence, vol. 8, no. 6, pp. 679-698, November 1986.Canny J. A computational approach to edge detection. IEEE Trans. Pattern Anal. Mach. Intell. 1986. Vol. 8, no. 6. Pp. 679-698S. Russell, P. Norvig, J. F. Canny, J. Malik, and D. D. Edwards, Artificial Intelligence: A Modern Approach, 2nd ed., Prentice Hall Series in Artificial Intelligence, Upper Saddle River, NJ: Prentice Hall/Pearson Education, 2003.Soinikova E.S., Ryabihk M.S., Batishchev D.S., Mikhelev V.M. High-performance method for boundary detection in medical images// Academic science – problems and achievements IX: Proceedings of the Conference. North Charleston, 20-21.06.2016—North Charleston, SC, USA:CreateSpace, 2016, p.93-95.NVIDIA, “NVIDIA CUDA C programming guide – version 7.0,” NVIDIA developer website, June 2016. [Online].  Available:  http://docs.nvidia.com/cuda/cuda-c-programming-guide/#axzz4IHtkC9CZ.