基于超像素稀疏表示的图像超分辨率方法研究
Project Number61461028
刘微容
Abstract超高清显示技术的快速发展必将对高空间分辨率的图像产生巨大需求,受成像工艺、成本和环境等因素制约,高空间分辨率的图像难以广泛获取,亟待研究能有效提升现有低质量图像空间分辨率的方法。虽然基于稀疏表示的图像超分辨率取得了一些研究成果,但是采用像素或图像块的稀疏表示方法对图像复杂结构表示不够准确,且过完备字典未能充分利用图像自身信息和样本图像先验信息。因此,基于稀疏表示的图像超分辨率方法仍有很大性能提升空间。本项目将深入研究基于超像素稀疏表示的图像超分辨率方法,包括提出图像复杂结构的超像素表示方法,提出内容和尺度自适应的高-低分辨率超像素字典学习方法,建立高-低分辨率超像素间的关系模型,提出基于超像素的结构化稀疏表示模型和求解算法。突破像素或图像块等方式对图像复杂结构超分辨率的限制,精确重构低分辨率图像缺失的细节信息,提升图像超分辨率方法的性能,在低分辨率壁画图像的高清展示中取得实际应用。【英文摘要】Withtherapiddevelopmentsofultra-highdefinitiondisplaytechnologiesinrecentyears,thedemandforhighspatialresolutionimagesgrowsfaster.Itisimportanttoreconstructthehighresolutionimagefromthecorrespondinglow-resolutionimageusingdigitalimageprocessingtechniques,withoutupdatingtheimagingequipment.Severalimagesuperresolutionmethodsbasedonsparserepresentationhavebeenproposed.However,thecomplexstructuresinanimagecannotbeaccuratelyrepresentedbythetraditionalpixel-levelorblock-levelsparserepresentation,andthepriorinformationinthelowresolutioninputimagehasnotbeenusedforover-completedictionarieslearning.Tosolvethoseproblems,thisprojectfocusesoninvestigatingimagesuperresolutionmethodviasparserepresentationwithsuperpixel.Combiningsparserepresentationtheorywithsuperpixelsegmentationmethodandmultiscaleanalysistheory,firstly,wewillproposeanewsuperpixelrepresentationmethodofcomplexstructuresinanimage.Secondly,wewillpresentanewsuperpixel-leveldictionarylearningalgorithmwithcontentandscaleadaptive.Thirdly,wewilldesignastructuralsparserepresentationmodelwithsuperpixel,andapriormodelbetweenthehigh-resolutionsuperpixelsamplesandlow-resolutionsuperpixelsamples.Finally,wewillproposeanewimagesuperresolutionmethodviasparserepresentationwithsuperpixeltoaccuratelyreconstructfinedetails,andimprovetherobustnessofimagesuperresolutionmethod.Theachievementscanbeappliedtosuperresolutionofthelow-resolutionimagesfrompaintedmurals.
Subtype地区科学基金项目
Project Source国家自然科学基金
2015-01
End Date2018-12
MOST Discipline Catalogue12 - 信息科学 ; 1201 - 电子学与信息系统
Host Institution兰州理工大学
Project Funding430000.0
CountryCN
Language中文
Document Type项目
Identifierhttp://ir.lut.edu.cn/handle/2XXMBERH/64703
Collection党委教师工作部(人事处、教师发展中心)
Recommended Citation
GB/T 7714
刘微容.基于超像素稀疏表示的图像超分辨率方法研究.2015.
Files in This Item:
There are no files associated with this item.
Related Services
Usage statistics
Google Scholar
Similar articles in Google Scholar
[刘微容]'s Articles
Baidu academic
Similar articles in Baidu academic
[刘微容]'s Articles
Bing Scholar
Similar articles in Bing Scholar
[刘微容]'s Articles
Terms of Use
No data!
Social Bookmark/Share
All comments (0)
No comment.
 

Items in the repository are protected by copyright, with all rights reserved, unless otherwise indicated.