电信科学 ›› 2015, Vol. 31 ›› Issue (4): 37-42.doi: 10.11959/j.issn.1000-0801.2015103

• 专题:视频技术与应用 • 上一篇    下一篇

分布式并行计算在超分辨率图像增强中的应用

郑杰1,包盛1,杨萤2   

  1. 1 百视通新媒体股份有限公司研究院 上海 200233
    2 上海文化广播影视集团有限公司 上海 200041
  • 出版日期:2015-04-15 发布日期:2015-04-15
  • 基金资助:
    上海市科学技术委员会“基于4K关键技术的互联网电视产业化应用示范”基金资助项目

Application of Distributed Parallel Computing in Super-Resolution Image Enbancement

Jie Zheng1,Sheng Bao1,Ying Yang2   

  1. 1 Research Institute of BesTV New Media Co.,Ltd.,Shanghai 200233,China
    2 Shanghai Media Group Co.,Ltd.,Shanghai 200041,China
  • Online:2015-04-15 Published:2015-04-15
  • Supported by:
    Internet TV Industrialization Demonstration Based on Key Technology of UltraHD

摘要:

主要介绍了分布式并行计算在超分辨率图像增强领域中的应用,超分辨率图像增强是指在不升级现有采集设备、不增加传输带宽的情况下,对现有1080P的视频内容通过软件方法提高图像分辨率,从而使图像达到4K的画面质量。由于超分辨率图像增强技术需要强大的计算能力支持,研究了分布式并行计算环境下的超分辨率图像增强系统的搭建以及性能对比测试,得出了分布式并行计算技术可以成百倍地提升超分辨率图像增强处理能力的结论,在实际应用中可大幅提升4K 内容的生产能力。

关键词: 超分辨率, 图像增强, 分布式计算, 并行计算

Abstract:

The distributed parallel computing in the super-resolution image enhancement applications was introduced.Super resolution image enhancement refers to improving the image resolution of 1080P video to 4K by software without upgrading existing acquisition device and increasing the transmission bandwidth.Due to the fact that super resolution image enhancement technology needs enormous computing power,super resolution image enhancement system and performance comparison testing by distributed parallel computing was studied.The conclusion shows that the super-image enhancement processing ability can be enhanced hundreds of times with distributed parallel computing technology,so that the production capacity of 4K video can be improved in rea1-world application.

Key words: super resolution, image enhancement, distributed computing, parallel computing

No Suggested Reading articles found!