基于压缩感知的高动态范围图像采集存储与重现方法的研究

项目名称： 基于压缩感知的高动态范围图像采集存储与重现方法的研究

项目编号： No.61473322

项目类型： 面上项目

立项/批准年度： 2015

项目学科： 其他

项目作者： 谭洪舟

作者单位： 中山大学

项目金额： 85万元

中文摘要： 数字化图像为人们获取外界信息提供了重要的途径。但传统的数字图像因动态范围太小而无法描述真实世界的光线变化，故高动态范围（High Dynamic Range,　HDR）图像技术应运而生。但因HDR图像容量太大，导致其在采集、存储、传输、处理、重现等5个重要技术环节，均存在未能较好解决的问题。为此，本项目提出基于压缩感知理论，解决HDR图像采集的根本问题，并解决与之相关的存储与重现问题，内容包括：（1）利用LDR采集设备采集HDR图像的研究；（2）基于压缩感知的HDR图像重建算法的研究；（3）基于压缩感知的HDR图像存储方法的研究；（4）基于HVS模型的HDR图像重现的研究；（5）硬件系统中实时的HDR图像的采集、存储与重现系统研究。 以上问题的研究和解决，将有助于HDR图像的采集、存储、传输、处理、重现等技术的进一步改善，同时也为其硬件实现提供理论的基础。

中文关键词： 数字图像处理；压缩感知；高动态范围图像；稀疏表示；图像融合

英文摘要： Abstract: Digital image is playing an important role in accessing information from real-world scenarios. Since the limited dynamic range of traditional images, it is difficult to depict details of dynamic light in the scene. High Dynamic Range Imaging (HDRI) emerged to solve the problem. However, HDR images consume too much storage space, which leads to the trouble of acquisition, storage, transfer, processing and representation. Our project is to solve the fundamental problems based on compressed sensing theory. The project includes: (1) the acquisition of HDRI using LDR(Low Dynamic Range) device; (2) the reconstruction of HDRI based on compressed sensing theory; (3) the storage of HDRI based on compressed sensing theory; (4) the representation of HDRI based on HVS(Human Visual System) model; (5) Hardware system of real-time HDRI acquisition, storage and representation. The research not only helps to improve acquisition, storage, transfer, processing and representation of HDRI, but also provide the fundamental theory for hardware design.

英文关键词： Digital Image Processing;Compressed Sensing;Image Registration