小波图像去噪算法在FPGA中的实现

分类号：密级：ＵＤＣ：编号：工学硕士学位论文小波图像去噪算法在FPGA中的实现硕士研究生：陈拓指导教师：卞红雨教授学科、专业：信号与信息处理学位论文主审人：王大成副教授哈尔滨工程大学2013年3月分类号：密级：ＵＤＣ：编号：工学硕士学位论文小波图像去噪算法在FPGA中的实现硕士研究生：陈拓指导教师：卞红雨教授学位级别：工学硕士学科、专业：信号与信息处理所在单位：水声工程学院论文提交日期：2013年3月论文答辩日期：2013年3月学位授予单位：哈尔滨工程大学ClassifiedIndex:U.D.C:ADissertationforDegreeofM.EngImplementationofWaveletImageDenoisingAlgorithmBasedonFPGACandidate:ChenTuoSupervisor:Prof.BianHongyuAcademicDegreeAppliedfor:MasterofEngineeringSpecialty:SignalandInformationProcessingDateofSubmission:Mar.2013DateofOralExamination:Mar.2013University:HarbinEngineeringUniversity哈尔滨工程大学学位论文原创性声明本人郑重声明：本论文的所有工作，是在导师的指导下，由作者本人独立完成的。有关观点、方法、数据和文献的引用已在文中指出，并与参考文献相对应。除文中已注明引用的内容外，本论文不包含任何其他个人或集体已经公开发表的作品成果。对本文的研究做出重要贡献的个人和集体，均已在文中以明确方式标明。本人完全意识到本声明的法律结果由本人承担。作者(签字)：日期：年月日哈尔滨工程大学学位论文授权使用声明本人完全了解学校保护知识产权的有关规定，即研究生在校攻读学位期间论文工作的知识产权属于哈尔滨工程大学。哈尔滨工程大学有权保留并向国家有关部门或机构送交论文的复印件。本人允许哈尔滨工程大学将论文的部分或全部内容编入有关数据库进行检索，可采用影印、缩印或扫描等复制手段保存和汇编本学位论文，可以公布论文的全部内容。同时本人保证毕业后结合学位论文研究课题再撰写的论文一律注明作者第一署名单位为哈尔滨工程大学。涉密学位论文待解密后适用本声明。本论文(□在授予学位后即可□在授予学位12个月后□解密后)由哈尔滨工程大学送交有关部门进行保存、汇编等。作者(签字)：导师(签字)：日期：年月日年月日哈尔滨工程大学硕士学位论文2小波图像去噪算法在FPGA中的实现摘要由于小波变换在图像处理等一系列领域内得到了越来越广泛的应用，利用小波变换对图像进行实时去噪的研究也日益成为业界的一个热点。随着电子技术的高速发展，以FPGA为首的现场可编程器件以体积小、并行度高、资源丰富及开发灵活等优势成为了新一代电子设计的主导方向，也为实现高速实时小波图像去噪算法提供了解决方案。本文采用Xilinx公司的Virtex-4-XC4VLX15芯片作为小波去噪算法的处理芯片，软件采用Xilinx的ISE开发工具，仿真工具采用ISE集成仿真工具和Modelsim相结合。开发高速小波图像去噪处理程序，实现小波图像去噪的实时处理。本文首先以小波理论作为时频分析手段，通过Matlab的计算和仿真，选取小波，对比常用去噪方法与小波去噪方法的效果，论证了小波去噪方法和其他去噪方法相比的优势，对比了不同去噪函数下对噪声抑制的效果，分析了噪声图像经不同去噪函数去噪后各个方面的差异；其次以5/3提升小波和对称延拓法为基础，设计出了一维小波变换逻辑架构、二维小波变换逻辑架构以及多层小波变换逻辑架构；再次综合Matlab仿真效果及计算量，选择小波去噪法中的通用阈值法作为去噪方法，完成了包括数据存储模块、阈值估计模块和去噪函数模块在内的去噪模块的整体逻辑架构设计；最后在由实验室自主开发的数字图像开发板中完成程序的硬件测试。通过软件仿真和硬件验证证明，小波图像去噪程序能够在硬件上稳定运行并达到预定的功能。关键词：小波变换；提升算法；小波阈值去噪；FPGA哈尔滨工程大学硕士学位论文小波图像去噪算法在FPGA中的实现ABSTRACTBecausethewavelettransforminarangeofareassuchasimageprocessing,hasbeenmorewidelyused.Real-timeimagedenoisingbasedonwavelettransformisalsoincreasinglybecomingahottopicofthescientificcommunity.Withtherapiddevelopmentofelectronictechnology,thefieldprogrammabledeviceledbyFPGAisbecomingthedominantdirectionofanewgenerationofelectronicdesign,becauseoftheadvantageofsmallsize,abundantresources,highparallellevelandflexibilityindesign.Alsoprovidesasolutionfortherealizationofhigh-speedreal-timewaveletimagedenoisingalgorithm.Inthispaper,IselectXC4VLX15FPGAofXilinxInc’sVirtex-4seriesaswaveletimagedenoisingalgorithmprocessingchip,anddevelopsoftwarebaseonISEofXilinxcompany,SimulationtoolsuseISEintegratedsimulationtoolsandModelsimsimulationtools,fordevelopinghigh-speedwaveletimagedenoisingprocessingprogramtoachievethereal-timeprocessingofwaveletimagedenoising.Inthispaper,themethodoftime-frequencyanalysisiswavelettheory.Firstofall,viaMatlabcalculationandsimulationtoselectwavelet,demonstratestheadvantageofwaveletdenoisingmethodthroughcomparetheresultofcommondenoisingmethodandwaveletdenoisingmethod.Analysisofallaspectsofthedifferencesoftheimageafterusingdifferentdenoisingfunctionbycontrastingthedifferentdenoisingfunctionundertheeffectofnoisesuppression.Secondly,designalogicalarchitecturewithone-dimensionalwavelettransform、two-dimensionalwavelettransformandmulti-levelwavelettransformbasedon5/3liftingwaveletandsymmetricextensionmethod.Thirdly,selecttheuniversalthresholdmethodasDenoisingtocompletethedenoisingmoduleincludingdatastoragemodule,thethresholdestimationmoduleanddenoisingfunctionmoduleoveralllogicalarchitecturedesignafterconsidertheeffectandcalculatetheamountoftheMatlabsimulation.Finally,completehardwaretestinthedigitalimageboardthatindependentlydevelopedbyourlaboratory.Waveletimagedenoisingprocessworkstablyandcouldachievetheintendedfunctionhardwarebycompletingsoftwaresimulationandhardwareverification.Keywords:WaveletTransform;LiftingScheme;WaveletThresholdDenoising;FPGA哈尔滨工程大学硕士学位论文小波图像去噪算法在FPGA中的实现目录第1章绪论·······························································································11.1立题背景························································································11.2相关技术研究现状············································································11.2.1小波及小波去噪方法的研究现状·················································11.2.2FPGA小波技术的研究现状·························································21.3论文的主要研究内容·········································································3第2章小波理论与去噪算法的研究·································································52.1小波变换基本理论············································································52.1.1连续小波变换··········································································52.1.2离散小波变换··········································································62.1.3多分辨率分析··········································································62.1.4图像小波变换··········································································72.2图像去噪质量评价············································································82.2.1主观评价方法··········································································82.2.2客观评价方法···