Abstract—In this paper, a design scheme for hardware acceleration for ZYNQ SoC programming and medical image processing using SDSoC development software is introduced and compared with traditional hardware language programming and Vivado HLS programming. In SDSoC, developers can use C/C++ language for hardware development as well as OpenCV and xfOpenCV library. OpenCV is a widely used image processing library, which can not only to shorten the development cycle and reduce the difficulty of hardware development, but also take up more FPGA resources. XfOpenCV is an OpenCV library optimized by Xilinx, similar in usage to OpenCV. In this paper, several typical images processing algorithms are used to test medical image data onto DICOM format, and the design of the proposed scheme and the traditional scheme is compared with the perspectives of processing speed comparison, power consumption, and development cycle. Sobel filter, Gaussian smoothing, and Harris corner detection are chosen for comparison study for their widely usages in image processing. Finally, performance on four platforms - CPU, ARM, ZYNQ and GPU are compared and evaluated to our method.

Index Terms—ZYNQ, SDSoC, medical image processing, hardware acceleration

Cite: Liang Mu, Tao Wei, Yuyu Tao, Chang Liang, and Xuejun Zhang, "Design of Medical Image Hardware Acceleration Platform by SDSoC for ZYNQ SoC," Journal of Image and Graphics, Vol. 8, No. 4, pp. 98-106, December 2020. doi: 10.18178/joig.8.4.98-106

Copyright © 2020 by the authors. This is an open access article distributed under the Creative Commons Attribution License (CC BY-NC-ND 4.0), which permits use, distribution and reproduction in any medium, provided that the article is properly cited, the use is non-commercial and no modifications or adaptations are made.