匀速直线运动模糊图像复原的改进算法
一篇很好的关于模糊图像复原的论文，含详细需要的函数，讲解清楚易懂。闫永存,等匀速直线运动模糊图像复原的改进算法级的概率密度函数P(s)如下式所示:P(s)=P(rdr(6)8对于连续图像,当直方图均衡化(并归一化)后有P(s)即ds=p (r)dr =dT(r)3两边取积分得x=7(=P)h(8)式(8)就是所求的变换函数对于离散图像,假定数字图像中的总像索为N,灰度级总图3最优窗法区域分布图数为L,第k个灰度级的值为r,图像中具有灰度级r的像素Fig 3 Areal distribution of optimal window method数目为n,则该图像中灰度级n的像素出现的概率为取值范围为V-P+1,V1-11;在水平方向上,区域1、2、3的横P(r)="O≤r≤1;h=0,1,…(9)坐标取值范围1O,P1-2],区域4.8、9的横坐标取值范围为P3-1,V1-P3m,区域5、67的横坐标取值范围为[V1-Psn+对其进行均匀化处理的变换函数为:VL-11在图3中,每一个区域都有各白独立的边界,即各个S=7(r)=∑P(n)=2N子窗区域的尺寸不一定相同(10)最优窗计算公式为利用式(10)对图像做灰度变换,即可得到直方图均衡化后的图像。∑∑h(p,g)∑∑h(p.q)∑∑(p,q)D=O 0该方法可以将滤除高频噪声,提高有用信号的嗝度,增加∑h(p∑∑h(P)(12)对比度,同时缩小叠加噪声信号的动态范围,抑制振铃效应有效的结合起来,高文硕等证明了这一点。但不能完全去除振220)2A)铃效应,因此文中在滤波前用最优窗法对图像进行处理最优窗对模樹图像的边界进行加杈处理,以致像素值向3.4最优窗法外逐步过渡到零,其目的是待处理图像的边界结合处不会出在恢复图像过程中,由于图像边缘的像素没有足够的相现灰度的跳变,振铃效应因此得到抑制。邻像素可以利川,所以会导致恢复图像的边缘变差,并且整幅图像有明暗相间的条纹,即振铃效应。为了解决这个问题,早4实验结果及分析期学者常采用边界修正法,但效果不够令人满意。 Aghasi在1996年提出循环边界法,其缺点是图像尺寸变为原来的4文中通过实验验证了改进算法的可行性和有效性,以sadhna模糊图像的复原为例,图4(a)是原始图像,对其进行倍,运算量增加很多。基于循坼边界法的缺点, Bimetal提出了对二维模糊图像四进行恢复的最优窗。其具体实胞过程为:模糊加噪运算,模糊角度为53°,模糊长度为45,高斯噪声为0.0l图4(b)是降质后的模糊图像,采用本文算法估计岀的模恢复窗a,k将图像平面分成9个区域,每个区域编号如图3糊方向为51°,模糊长度为46,图4(c)是普通维纳滤波复原所示。标号为9的中央区域o:=1。图像,图4(d)是人工调整参数为真实值的复原结果。利用本区域1,8、7的纵坐标取值范围为[0,P-2],区城2,6、9文的改进算法得到的复原结果如图4()所示。实验结果如图的纵坐标取值范围为P-1,V4-P,区域3、4.5的纵坐标4所示。(a)原始图像(b)模糠图象(c)普通维纳波复原图象d)取冥际参数值的复原图像(e)太文算法的复原图(a)original image(t)blurred image (c)restored image of ordinary (d)res tored image fcr the (e)restored image ofiner filteringactua l )arameter valles algorithm in this pa per图4运动模糊图像及复原结杲Fig. 4 Motion blurred images and results of restoration147C1994-2012ChinaAcademicJournalElcctronicPublishingHousc.Allrightsrescrvcd.http://www.cnki.net电子设计工程》2012年第3期出实验结果可知,方向徵分法可以很人程度地提高模糊2005,10(5):590-595角度的估计准确性,利用自相关函效负尖峰值可以较准确地「31贺卫国,黎绍发,匀速直线运动模糊长度的精硝估计[鉴别出模糊长度,从而可以提高图像还原质量。最优窗法对振计算机应用,2005,25(6):1316-1320铃效应可以有较好的抑制作用,最后得到了复原效果较为理HIE Wei-guo, LI Shao-fa. Estimating the blurring length of想的图像。uniform linearmotion blurred images[J]. Computer Applications5结束语2005,25(6:1316-1320[4]吴振字.模糊图像复原方法研究[D]长沙:国防科学技术文中对运动模糊图像的退化模型、维纳滤波复原原理、点大学,2009扩散函数的求取过程选行了详细削述,提出了一种改进的模[5]高文硕.郊伟伟,杨磊运动模糊图像复原技术的改进算法糊图像复原算法,并对振铃效应进行处理,以 sadhna图像的中国传媒大学学报自然科学版,2010,7(1):72-76复原为例进行了实验验证结果表明,文中方法可以较准确地GAO Wen-shuo, ZHENG Wei-wei, YANG Lei. Improved估计出运动模糊参数,并且提高了运算速度,振铃效应得到有algorithm for restoration of the imagemotion blur IJ]. Journal效抑制。of Communication L niversity of China Science and Technology参考文献2010,17(1):72-76[1] Cannon M. Blind deconvolution of spatially invariant image [ 6] Aghdasi F, Ward R K. Reduction of boundary artifacts inblurs with phase [ J]. IEEE Trans on Acoustics, Speech andimage restoration[J]. IEEE Trans. Image. Proc. 1996. 5(4)signal Processing, 1976(24): 58-63611-6182]陈前荣,陆启生,成礼智,基于方向微分的迈动模糊方向鬥叶海.基于统计特征加权的模糊聚类方法及其应用鉴别!中国图象图形学报,2005,10(5):590-595现代电子技术,2009(1299-102chen Qiall-rong, LU Qi-sheng, CHENG Li-zhi. IdentificalionYE Hai-jun. Fuzzy clustering method and its applicationof motion blur direction from motion blurred image bybased on statistical characteristics weighting [J]. Moderndirection derivation method [J Journal of Image and GraphicsElectronics Technique, 2009(12): 99-102具ⅢP2优化和DC偏移消除的宽带丨/Q解调器可改善接收器性能加利福尼亚州米尔皮塔斯( MILPITAS,CA)推出超宽带宽直接转換lQ解调器LIC∶5S35,该件具卓越的线性性能(在1.95GIlκ时,ⅢP3-25.7dBm,IP2-60dBm)。LTC585能提供超过530MILz的基带输岀解调带宽,可满足新-代霓带LTE多模式接收器和毅宇预失真(DPD)接收器的带宽需。Q解调器在700WHx至3GHz的宽频率范围内二作,几乎覆盖了所有蜂窝基站频毁。这款器伫的独狩之处是两个内置的校准功能。其一是允许系统设计人员优化接收器IP2性能的高级也路以60dBm标称值提升至前所未有的80dBm或更高。另一个则是用于消除I和Q输出端上的DC偏移电压的片内电路。这两个功能电路均起到了増强接收嚣性能的作用。此外,LTC5585还可提供超卓的16 dBm pldB。为了进一步加强该器件在直接转换接收器应用中的使用,LTC5585捉供非常低的IQ幅度和相位失配ε幅度失配的典型值是0.05dB,而相位误差的典型值是0.7度,两个数值都是在1.95GHz频率上测得的。这两者的结合产生了一个43dB的接收器镜頰抑制能力。因为LTC5585能湜供非常宽的带宽,所以尢其适用于多模式LTE、 W-CDMA和TD- SCDMA基站DPD接收器以及于主收器的应門。尤其是对;DPD,这些最新一代基站正在将解调带宽推进到超过300MHz。LTC5585可以非常方便地配置应对这些带宽的挑战。除了无线基础设施应用,LrC5585还适用于军用接收器、宽蒂通信、点对点微波数据链路、镜频抑制接收器和长距离RFID阅读器。LTC5585内置了一个RF变压器以减少外部组件,再加上24引线4mm×4 mm QFN封装,因而可提供高度紧凑的解决方案。该器件规格在_40-105℃C的外壳工作温度范围。LTC585用单—5Ⅴ电源供电,吸取200mA的总电源电流。该器件提供数竽输λ以启用或停用该芯片。当俘用时,该IC吸取的典型溻电流为11μA。解调器的200ms快谜接通时间和800rs断开时间使该器件能在突发模弌接收器中使用。咨询编号:2012031009心·心;心·心,心·心普心心心·心·心心·心分·心心···心·心心心心·心·心心·心欢迎订阅2012年度《电子设计工程》(半月刊)国内邮发代号:52-142国际发行代号:M2996订价:15.00元/期360.00元/年148C1994-2012ChinaAcademicJournalElcctronicPublishingHousc.Allrightsrescrvcd.http://www.cnki.net
- 2020-11-27下载
- 积分：1
Adaptive Filters
自适应滤波器的权威Sayed的大作，用大量的实例诠释自适应滤波器的各种算法原理，是不可多得的好书。ADAPTIVE FILTERSADAPTIVE FILTERSALIH SAYEDUniversity of California at Los Angeles◆旧EEEIEEE PressWIlEYNTERSCIENCEA JOHN WILEY SONS, INC, PUBLICATIONCover design by Michael RutkowskiCopyright C 2008 by John Wiley Sons, Inc. All rights reservedPublished by John Wiley Sons, Inc, Hoboken, New JerseyPublished simultaneously in CanadaNo part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or bymeans, electronic, mechanical, photocopying, recording, scanning, or otherwise, except as permitted under anySection 107 or 108 of the 1976 United States Copyright Act, without either the prior written permission of thePublisher, or authorization through payment of the appropriate per-copy fee to the Copyright Clearance Center,Inc, 222 Rosewood Drive, Danvers, MA01923, (978)750-8400, fax(978)750-4470, or on the web atwww.copyright.comRequeststothePublisherforpermissionshouldbeaddressedtothePermissionsDepartment, John Wiley sons, Inc, 111 River Street, Hoboken, NJ,(201)748-6011, fax(201)748-6008,oronlineathttp:/www.wiley.com/go/permissionLimit of Liability/Disclaimer of Warranty: While the publisher and author have used their best efforts inpreparing this book, they make no representations or warranties with respect to the accuracy or completeness ofe contents of this book and specifically disclaim any implied warranties of merchantability or fitness for aarticular purpose. No warranty may be created or extended by sales representatives or written sales materialsThe advice and strategies contained herein may not be suitable for your situation. You should consult with aprofessional where appropriate. Neither the publisher nor author shall be liable for any loss of profit or anyother commercial damages, including but not limited to special, incidental, consequential, or other damagesFor general information on our other products and services or for technical support, please contact our CustomerCare Department within the United States at(800)762-2974, outside the United States at (317)572-3993 or fax(317)572-4002Wiley also publishes its books in a variety of electronic formats. Some content that appears in print may not beavailable in electronic formats. For more information about wiley products, visit our web site atwww.wlley.conLibrary of Congress Cataloging-in-Publication Data:Sayed, Ali hAdaptive filters/Ali H. Sayedp cmIncludes bibliographical references and indexISBN9780470-25388-5( cloth)Adaptive filters. I. TitleTK7872F5s285200862138l5324dc222008003731Printed in the united states of america10987654321To my parentsContentsPrefacexviiNotationXXVAcknowledgmentsXXXBACKGROUND MATERIALA Random variablesA 1 Variance of a random variableA2 Dependent random VariablesA3 Complex-Valued Random VariablesA 4 Vector -Valued Random variables3467A.5 Gaussian Random VectorsB Linear Algebra12B. Hermitian and Positive- Definite matricesB 2 Range spaces and nullspace of matrices14B3 Schur Complements16B 4 Cholesky factorizationB 5 QR DecompositionB6 Singular Value Decomposition20B 7 Kronecker productsC Complex Gradients25C 1 Cauchy-Riemann Conditions5C2 Scalar arguments26C3 Vector arguments26PART:。 PTIMAL ESTIMATIONScalar- Valued Data291. 1 Estimation Without observations1.2 Estimation Given Dependent observations1.3 Orthogonality Principl36CONTENTS1,4 Gaussian random variables382 Vector- Valued Data422. 1 Optimal Estimator in the vector Case422.2 Spherically Invariant Gaussian Variables462. 3 Equivalent Optimization Criterion49Summary and Notes51Problems and Computer Projects54PART I: LINEAR ESTIMATION3 Normal Equatlons603. 1 Mean-Square Error Criterion613.2 Minimization by Differentiation3.3 Minimization by Completion-of-Squares633.4 Minimization of the error Covariance matrix653.5 Optimal Linear estimator4 Orthogonality princlple4. 1 Design Examples4.2 Orthogonality Condition4.3 Existence of solutions744, 4 Nonzero-Mean variables5 Linear Models5.1 Estimation using Linear Relations5.2 Application: Channel Estimation5Application: Block Data Estimation815. 4 Application: Linear Channel equalization825.5 Application: Multiple-Antenna Receivers85Constralned estimation876.1 Minimum-Variance Unbiased estimation6.2 Example: Mean Estimation6.3 Application: Channel and Noise Estimation916.4 Application: Decision Feedback Equalization6.5 Application: Antenna Beamforming1017 Kalman Filter1047.1 Innovations process7.2 State-Space Model106
- 2020-12-09下载
- 积分：1
