• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文期刊>Journal of Low Power Electronics >A Low Power, Scalable and Runtime Customizable Microprocessor Architecture for Image Processing
【24h】

A Low Power, Scalable and Runtime Customizable Microprocessor Architecture for Image Processing

机译：用于图像处理的低功耗，可扩展和运行时可自定义的微处理器架构

获取原文
获取原文并翻译 | 示例
           
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

A very important rule into communication is played by images. As consequence the commercial value of portable devices has been growth if they integrate electronic sensors for image acquisition and the availability of user-friendly image-editing software. The speed in extracting and processing information coupled to the ability to enhance the quality of a digital image with easy procedures, represents the challenge of combining hardware (microprocessors) and software (algorithms) partitioning in VLSI design for image processing. Moreover, the low-power design methodology and the area size are the right supervisor guides to maintain acceptable the time-to-use of the modern image processing devices and their own portability. In this context, this paper exposes an easily scalable and runtime customizable microprocessor architecture with a structure that combines the paradigm of vectorial processing and Very Long Instruction Word (VLIW) with high utilization of parallel resources on both data and instruction level. The microprocessor has shown no significant area increase, compared with the ST200 VLIW architecture which is based onto, while its performance is improved by 2.4 up to 4 times. Moreover the solution proposed implies significant power savings (about 32.4%) due to an instruction level power control mechanism, to the drastic reduction of code size and a massive use of well known low power techniques. For this reason it represents a possible answer for low energy dissipation, executing image processing applications.
机译：图像播放了一个非常重要的通信规则。结果，如果将电子传感器集成用于图像采集的电子传感器以及用户友好的图像编辑软件的可用性，则便携式设备的商业价值已经增长。提取和处理信息的速度耦合到增强具有简单过程的增强数字图像的质量的能力，代表了组合硬件（微处理器）和软件（算法）划分的挑战，用于图像处理的VLSI设计。此外，低功耗设计方法和面积大小是正确的主管指南，以便可接受现代图像处理设备的使用时间和自身的便携性。在此上下文中，本文公开了一种易于可扩展和运行时可自定义的微处理器架构，该结构将矢量处理和非常长的指令字（VLIW）的范例结合在于数据和指令级别的高利用率。与基于ST200 VLIW架构的ST200 VLIW架构相比，微处理器没有显着的面积增加，而其性能高达2.4次，最多4次。此外，提出的解决方案由于指示水平功率控制机制而言，提出了显着的功率节省（约32.4％），以急剧降低代码尺寸和众所周知的低功率技术的大规模使用。出于这个原因，它代表了低能量耗散，执行图像处理应用的可能答案。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱：kefu@zhangqiaokeyan.com

京公网安备：11010802029741号 ICP备案号：京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有

  • 客服微信

  • 服务号