摘要
目的:为了进一步探讨医学图像的压缩与高速传输的问题,本文在NIOSⅡ软核处理器下进行医学图像压缩与传输的软硬件协同设计的研究。方法:采用Altera公司开发的基于SOPC技术的NIOSⅡ嵌入式处理器,完成一个医学图像压缩以及高速传输系统。整个系统硬件部分主要包括大容量FPGA芯片、Cypress公司的EZ-USB FX2 Cy7C68013芯片以及外围存储器。采用适合FPGA实现的行列分解的方法,将复杂度高、消耗时间多的2D-DCT部分实做成IP核形式即硬件实现,并根据硬件电路实现的特点进行优化;通过NIOSⅡ定制指令的方法来调用此硬件加速部分提高整个算法的运行速度;采用USB2.0接口来完成图像数据的高速传输。结果:通过对照,我们发现系统取得良好的压缩效果,基本上不影响医学诊断。结论:该系统性价比高且能在短时间内以较低的成本实现,其使用新的技术新的方法给医学仪器设计领域带来一个新的亮点。
Objective： To further the study of the medical image compression and high-speed transmission, we have carried out the research of the software and hardware co-design under NIOS Ⅱ soft-core processor for medical image compression and transmission. Methods： Using NIOS Ⅱ Embedded Processor developed by Altera and based on the technology of SOPC, we realized a medical image compression and high-speed transmission system. The hardware mainly consists of large-capacity FPGA chip, Cypress＇s EZ-USB FX2 Cy7C68013 chip and external memory. With decomposition ranks method which is suitable for realizing in FPGA, the complicated, time consuming 2D-DCT is brought into a IP core that is realized in hardware, and in accordance with the hardware circuit characteristics of the optimization; With method of customization instructions in NIOS Ⅱ, the operating speed of calling algorithm is enhanced; USB2.0 interface used to complete the high-speed image data transmission. Results： By contrast, we found that the system has made a good compression, and basically not affect medical diagnosis. Conclusion： The system can fetch high cost performance in a short period of time to achieve lower costs, its use of new technologies and new methods to the design of medical devices brought a new spot.
出处
《中国医学物理学杂志》
CSCD
2008年第2期582-585,共4页
Chinese Journal of Medical Physics
基金
广东省科技社会发展计划(No.2004B10401012)
