Real-time image super resolution using an FPGA下载

weixin_39822095 2019-08-13 07:00:13

Image super resolution is the process of combining a set of
overlapping low-resolution images to produce a single highresolution
image. In this paper, a novel real-time superresolution
system is presented which is based on a weighted
mean super-resolution algorithm combined with the existing
fast an
相关下载链接：//download.csdn.net/download/kongxin0513/4857177?utm_source=bbsseo

...全文

29 回复打赏收藏转发到动态举报

写回复

回复

切换为时间正序

请发表友善的回复…

发表回复

Image super resolution is the process of combining a set of overlapping low-resolution images to produce a single highresolution image. In this paper, a novel real-time superresolution system is presented which is based on a weighted mean super-resolution algorithm combined with the existing fast and robust multi-frame super-resolution algorithm. The resource requirements of the proposed architecture scale linearly with the targeted image quality, making it ideally suited for a variety of real-time applications such as HDTV. Simulation results demonstrate a speed-up of three orders of magnitude over optimized software implementations with negligible loss to image quality.

基于FPGA的嵌入式图像处理系统设计，英文版，非扫描版，内容清晰。作者简介 Donald G Bailey is Associate Professor in the Institute of Information Sciences and Technology at Massey University, where he leads the Image and Signal Processing Research Group. His research interests include most aspects of image analysis, but in particular the algorithm development process, and training. Bailey has developed a Vision Image Processing System package which has been used in a wide range of image analysis applications. Current and recent projects include: image processing using FPGAs, real time produce grading using machine vision, super-resolution, and sub pixel measurement techniques, camera calibration, and coastal monitoring using automated video analysis. He has been working as an electronics and computer systems engineer in the field of image analysis and machine vision for over 25 years. He began applying FPGA technology to image processing in 2002, and since then has published about 25 papers on issues and applications of FPGAs to image processing. 目录 Preface. Acknowledgements. 1 Image Processing. 1.1 Basic Definitions. 1.2 Image Formation. 1.3 Image Processing Operations. 1.4 Example Application. 1.5 Real-Time Image Processing. 1.6 Embedded Image Processing. 1.7 Serial Processing. 1.8 Parallelism. 1.9 Hardware Image Processing Systems. 2 Field Programmable Gate Arrays. 2.1 Programmable Logic. 2.2 FPGAs and Image Processing. 2.3 Inside an FPGA. 2.4 FPGA Families and Features. 2.5 Choosing an FPGA or Development Board. 3 Languages. 3.1 Hardware Description Languages. 3.2 Software-Based Languages. 3.3 Visual Languages. 3.4 Summary. 4 Design Process. 4.1 Problem Specification. 4.2 Algorithm Development. 4.3 Architecture Selection. 4.4 System Implementation. 4.5 Designing for Tuning and Debugging. 5 Mapping Techniques. 5.1 Timing Constraints. 5.2 Memory Bandwidth Constraints. 5.3 Resource Constraints. 5.4 Computational Techniques. 5.5 Summary. 6 Point Operations. 6.1 Point Operations on a Single Image. 6.

开个新坑，最近在读FPGA+DL方面的论文，进行分类整理一下，暂时分为下面几类，如果有发现其他方面再慢慢添加。另外，这里先暂时只做分类整理，都是粗略的读了abstract和contribution，将每个方向的论文尽可能的找到源头，方便后续详读的时候按顺序进行。注意！！！如果有的论文链接打不开请及时留言如果有任何觉得缺失的欢迎私信，我会加进去快速计算快速计算分为两种，一种是利用快速算法，快速算法主要是FFT和Winograd算法，能够加快卷积运算。其中FFT更适合于卷积核较大的情况下，Winogr

SR_简单总结 ① GANrecursive networkResidual networkCNNAUTO ENCODERELSE SR_简单总结 ① ) GAN GANISRGAN: Improved Super-Resolution Using Generative Adversarial Networks （CVC) G-GANISR: Gradual generative adversa...

Visual Grounding with Transformer[206]具有编码器-解码器架构，其中视觉标记(从预训练的CNN模型中提取的特征)和文本标记(通过RNN模块解析)在编码器的两个不同分支中并行处理，具有跨模态注意力，以生成文本引导的视觉特征。尽管有人可能认为像Transformer模型这样的架构应该是通用的，可以直接应用于各个领域，但我们注意到，预训练这些模型所需的高计算成本和时间成本需要新的设计策略，使它们在视觉问题上的训练更具成本效益。

下载资源悬赏专区

13,655

社区成员

12,618,968

社区内容

发帖

与我相关

我的任务

其他技术论坛（原bbs）

社区管理员

加入社区

近7日
近30日
至今

加载中

查看更多榜单

社区公告

暂无公告

试试用AI创作助手写篇文章吧

+ 用AI写文章