Gabor变换重构的严重问题
对数据进行Gabor变换，用反变换进行重构，发现结果与原信号相差甚远Gabor正变换用的是时频工具箱里的tfrgabor程序，反变换程序是我自己编的
比如
y=log(load('data1.m')) %读数据
=tfrgabor(y); %正变换
=size(ytfr);
%反变换
y1=zeros(1,length(y));
for k=1:length(y1)
a=0;
for i=1:m
for j=1:n
a=ytfr(i,j)*(1/(2*sqrt(pi)*(2*m+2)))*exp(-(((k-i*length(y)/m))^2)/((2*m+2)^2))*exp(sqrt(-1)*2*pi*j*k/n)+a;
end
end
y1(k)=a;
end
figure(10)
plot((y1))
[ 本帖最后由 JulianChin 于 2007-11-30 03:15 编辑 ] 结果是……
[ 本帖最后由 JulianChin 于 2007-11-30 03:14 编辑 ]
data1.m全体数据
0.006648573485019 0.006245822249335 0.005843024842998 0.0054400917215890.005036892764129 0.004633260071229 0.004228990765071 0.003823849792423
0.003417572732058 0.00300986860824 0.002600422712089 0.002188899432857 0.001774945101308 0.001358190847513 0.00093825547553 0.000514748357542 0.000087272350068 -0.000344573265009 -0.00078118981187 -0.001222976221827 -0.001670326001267 -0.002123624163685 -0.002583244123089 -0.003049544546098 -0.003522866160132 -0.004003528515223 -0.004491826697052 -0.004988027989027 -0.005492368481309 -0.006005049624978 -0.006526234729671 -0.007056045403292 -0.007594557932667 -0.00814179960431 -0.008697744964744 -0.009262312020221 -0.009835358375988 -0.010416677315682 -0.011005993821837 -0.011602960538951 -0.01220715368101 -0.012818068885883 -0.01343511701954 -0.014057619933582 -0.014684806180187 -0.042261669159764 -0.015949650412528 -0.020934962761439 -0.028787128837749 -0.016694154207199 -0.027527159498321 -0.012032782786276 0.010962916916861 0.005697625760712 0.025101415449036 0.019657539179578 0.031382182020121 0.026368057853472 0.027898592478191 0.027477744924604 0.028901808066964 0.02532925444001 0.005999520012357 -0.000137053196484 0.013741496757913 0.012435602696573 0.008775842482736 0.008091336390792 0.00777873892512 0.003916826744558 -0.005749320476815 -0.005655690316632 -0.005450911419182 -0.014608346856584 -0.011527933910994 -0.00845229964172 0.004416160486922 -0.001830022790241 0.011806600397159 0.010046845271062 -0.000051996412167 0.004778190953679 0.00002713682288 -0.014960598309593 -0.004110233924365 0.001488625237092 -0.016067317765474 0.001858595283351 0.010992332575261 0.016615070364082 0.015060962272599 0.011655245922896 0.011756592735813 0.004192771665281 0.003043822153685 0.003495760404613 -0.001458302791374 0.0103235904531 0.010562701956836 0.011784593666117 0.01085310140627 0.004906156992336 0.006497223815801 -0.001389168089716 0.00064875546293 0.003206140333899 -0.005597202141135 0.006742806472348 0.006908965356359 -0.001188227986202 -0.000404326432347 0.000647417280876 0.002387801417905 -0.006716616912044 -0.005356406266366 -0.002015745812866 -0.010175354397471 -0.007890007816191 0.529302555487139 -0.008394429257994 -0.017938991501476 -0.020256030784174 -0.025266341552228 -0.017486874698439 -0.018910402649587 -0.018307912564129 -0.017118687190009 -0.018979681991184 -0.018915470556166 -0.026987086723595 -0.028428445301818 -0.026117430087749 -0.028732033677165 -0.019636628305237 -0.012357677182052 -0.012273823819381 -0.01234221290971 -0.010995756847862 -0.026963368716241 -0.032338448902514 -0.03764185484765 -0.037557698964084 -0.040394381280882 -0.046257025759809 -0.024906692037817 -0.026838802604856 -0.026819987379137 -0.033251586918264 -0.030146641010887 -0.044916911284163 -0.034891394650881 -0.022879563710343 -0.019766284343688 -0.024822828732064 0.010984377733708 0.015509562690196 0.016968296741071 0.016334412363346 0.009105846232299 0.001908417520507 0.005532936818181 0.012981265603987 0.014496856125198 0.013429506283518 0.013050916129492 0.014330865532458 0.012945688207854 0.007909960949355 0.00163945659304 -0.003817484750313 -0.002422586326797 0.01041702470197 0.010714854656412 0.002548746636191 -0.000685113348339 0.004911755375572 0.002329491242967 0.005618430130059 0.006832038149416 0.006340123587908 0.006672527300615 0.002171035624709 -0.003800854301886 -0.005382701880624 -0.004500024000686 -0.006342177657477 0.000147184949076 0.000516376619574 0.005449013262735 0.007350770331264 0.000795057463578 0.002231594436193 -0.002635417377871 0.003173404566244 0.007348177105809 0.009276424038697 0.007647523602248 0.01172153573549 0.010867259075895 0.002795068594199 0.002224730741137 0.003888987735109 0.002256760123298 0.001100200234348 0.005751462717665 0.007759016733352 0.004475790720135 -0.003639471583623 -0.000086721639253 -0.000138540859381 -0.008508895528911 -0.003921362880451 0.003337734656375 0.007848406078714 0.008716903172343 0.007270094776594 0.00653081684465 0.007959403303643 0.007880112311367 0.006567382828708 0.005154318578951 -0.0001739806594 -0.001692547260247 -0.002509185731972 0.003902946064732 0.002905788408087 0.00179203454514 -0.008371116981229 -0.005008304447648 -0.003510223594569 0.00504005190192 0.006221338685648 0.006998374731168 0.016707738726543 0.012994988191964 0.01074100619293 0.0117486402339 0.006898918844359 0.006997597400315 0.007576762133175 0.007500996520415 0.012495041270962 0.011332525807374 0.010643288446015 0.00251213212938 0.006451726443835 0.005789382262909 -0.003851753491372 0.000004173527762 -0.000517663532766 0.000688759035885 -0.005405756540743 -0.012788702417106 -0.008927956073252 -0.002128706088383 -0.005940994827045 -0.002385669788044 -0.002002716790947 -0.003358010311445 -0.000638182376576 -0.003055907607739 -0.012100763532613 -0.011244968308067 -0.013702402039292 -0.007581383531434 0.003657533864532 0.003235767694359 0.004507771692432 0.000266380386707 0.001854871507643 0.003090701619457 0.002384363144614 -0.001572302624194 0.000354850910695 -0.000856151108075 0.004784565123547 0.004905802755419 0.009685652583365 0.007059592210789 0.006538782741931 -0.000055878518894 -0.012644632280273 -0.01281383101676 -0.007853823694902 -0.011532197236233 -0.006259952510004 -0.002005086434878 0.001663955489326 0.003205953925185 0.002663950620028 -0.00254841508238 0.002701299622292 0.004338323318601 -0.006502372066007 -0.001743524542413 -0.000746473744987 -0.006461352389596 -0.022396935891624 -0.014775666485484 -0.005776919878882 -0.02371107946826 -0.025212716145677 -0.00548885632115 -0.018185017105322 -0.004828342530445 0.006450032224527 0.002616617935164 0.001360114984091 0.008312108824057 0.005967261075099 0.013356288334594 0.013437527141155 0.005652230768187 0.007768467158638 -0.013583391103735 -0.019104505535625 -0.010060415782192 -0.015031917313017 -0.014294679661868 -0.009188366690837 -0.019092282265485 -0.019317483836959 -0.010438990134408 -0.007056342735445 -0.028400314015925 -0.03393570485712 -0.02398958955594 -0.011624675284168 -0.007576691438129 -0.009453527961701 -0.003824789023565 -0.003188404226663 0.014251033229125 0.011763451016081 -0.001012176402438 -0.009681051514529 0.005741410637098 -0.011855851245049 -0.00114724225885 -0.013057744616753 -0.01525842453044 -0.011505282206191 -0.008199406340411 -0.012944534425369 -0.00683796160039 0.009535183489557 0.010800858117067 0.007681656343714 0.018398549208237 0.030703085521991 0.031869972150998 0.030959972713903 0.01976547481039 0.018572293702637 0.014780668844667 0.011489226516691 0.015914701565691 0.017899808608693 0.017623062147197 0.016063469348534 0.013951177796061 -0.003612932585528 -0.000994226042343 -0.001297350425643 -0.00235387127784 -0.003815622046796 0.003715285239229 0.002654896619874 0.012863856178968 0.0146221549678 0.010178886175737 0.003823702389831 0.001422515084738 0.005535922431611 0.00751582471176 0.000048345728273 0.006504376098843 0.007552738380179 0.006611340725092 0.009260181137725 0.003342261743879 0.009480826805412 0.011268045514019 0.009943980688416 0.008337754756965 0.004260613354983 -0.005880448340716 -0.007649734438511 -0.014737370908806 -0.009594157864266 -0.011356581889491 -0.009212901410291 0.001184472385859 -0.000135293393861 0.00181364554543 -0.00607951696509 0.004722708995462 0.001451167586684 0.00054948175609 0.004714249197734 -0.002369749414829 0.001302634365762 0.001717147376955 0.003802591415675 -0.003065051541904 -0.00612063728497 -0.01565219605796 -0.01192782506801 -0.013997413378323 -0.012043992529636 -0.020187200160364 -0.01156965748837 -0.011340965514841 -0.009113365364938 -0.011002447273111 -0.01232588097185 0.006701155385097 0.001693013435755 0.006147907450953 0.010349393828073 0.01092029376571 0.010297694511728 0.017184871453626 0.014314073801442 0.015286806895675 0.012264763224771 0.003548149690978 0.004182477472416 -0.01511962623559 -0.009549465911389 -0.005882636929985 -0.006486860891268 0.003814821956241 0.004527389382075 0.004824368548745 0.001345221511989 -0.000425513466506 0.000565189562965 0.007183669305017 0.01080787381601 0.013004409379118 0.009614001932285 0.004874945643836 0.004841259402415 0.002849955696663 0.004890639058453 0.010742445583379 0.010304189115352 0.013512936777248 0.014036454058481 0.002804788750704 0.010042693707735 0.010885118007817 0.020289354379438 0.019066321127613 0.019926587204882 0.017318286697706 0.013102153985523 0.007893745458659 0.005147512881911 -0.003386076666302 -0.004348544157668 0.000331389096931 0.001902165570248 -0.000361625787649 -0.008003096439127 -0.00894117360728 -0.004753215467941 -0.006229813929187 -0.005707648893506 -0.00930140501019 -0.018593789953882 -0.014458827164169 -0.02103057449504 -0.036895676339876 -0.02584745647848 -0.021039804056434 -0.018215635172856 -0.034432869972699 -0.035564088304467 -0.020076275467875 -0.034600288216009 -0.024211648019561 -0.026926225132186 -0.034935715217042 -0.034028418257274 -0.027777483917108 -0.025567312310524 -0.000866384156866 0.008046190383262 -0.000719855757479 -0.001690750669599 -0.0016946849694 -0.001971600146989 -0.002536448461527 -0.006209931847205 0.002912500396192 0.010755529528572 0.016276343920718 0.007950830989317 0.007412661191219 0.010621341213203 0.018114072469046 0.018773167405934 0.010757605449478 0.014476965061575 0.011992768757813 0.002303624447467 -0.000694997069742 0.002164698546972 0.001071526168773 -0.008798860086946 -0.005648395348598 -0.011627230436547 -0.00564172479264 -0.007955968309039 -0.001370943956045 0.002029146747898 0.00039394509995 -0.000137365925721 0.01202163670768 0.000243748173878 0.012175862008937 0.0094571203478 0.015595586015774 0.008705698296507 0.010703808802047 0.004173436994167 0.007075027224253 0.014982171124666 0.022646774565539 0.022196686180597 0.01543708510096 0.014829623778317 0.022895439882223 0.017699148545717 0.013553301086431 0.012263416921319 0.016506053577175 0.011599656329974 0.018741684268491 0.014531029505801 0.008380730059491 0.014505376179552 0.013947920561906 0.011945849018892 0.011698887977686 0.010082680954367 0.004515167239614 -0.006414691367893 -0.005520603636294 -0.004816042355451 -0.004102045480343 -0.003379953083946 -0.00265107675281 -0.001916704601588 -0.001178106334678 -0.000436538346026 0.000306751151719 0.001050516977419 0.001793511819724 0.00253448107202 0.003272157672749 0.019615887539964 -0.002664761635032 -0.001944875353108 -0.001233742478754 -0.000532774607472 0.007553858415634 -0.014563007270577 -0.013900993933447 -0.013254750393994 -0.012625530770911 -0.012014497136013 -0.011422723883823 -0.010851201974855 -0.010300843045151 -0.009772483375186 -0.009266887711799 -0.008784752937362 -0.008326711580856 -0.007893335166034 -0.007485137392266 0.008293487840808 -0.010930328959991 -0.010600322069946 -0.010297212701036 -0.010021315832939 -0.009772874936911 -0.009552063891742 -0.00517472107498 -0.009193689066604 0.001706986036108 -0.017013322730937 -0.028783144421316 -0.017282924981602 -0.021276031121607 -0.012320324725301 -0.020970496854664 -0.01424780032313 -0.033866742974741 -0.02062158096162 -0.026658483465869 -0.029736357597554 0.008847973302076 -0.003162020640321 0.008474694256847 0.002735917510296 0.002505856637487 0.010653177709738 0.004457917138712 0.001941967064017 -0.002785289603915 -0.005771183306537 0.000481945339425 0.000909549992617 -0.013036126837463 -0.005676909363184 -0.008378553521408 -0.008637245490996 -0.007942652796581 -0.00888682212741 0.001126862161702 0.005850170336762 0.006111218747285 0.004117440320885 0.012055151653468 0.019449488974675 0.013703353309919 0.021200435777055 0.019255004243745 0.029426005961009 0.034196234727702 0.031576485864086 0.022013129270114 0.021734310155322 0.022079207999349 0.016538374953375 0.023103589845094 0.013999211143426 0.006324533974853 0.001000206744352 0.015806406831716 0.016079054409542 0.015679711070428 0.023961381893195 0.01004323848336 0.007651685474701 0.01082345791177 0.008075988298515 0.001656507539758 -0.006299256410058 -0.003090724473943 -0.002991647304285 -0.007298446907409 -0.010752221841869 -0.006659824179564 0.000348925877028 0.008749765130743 0.018885782420689 0.009870325635315 0.010494386413532 0.005141987105405 0.014452599785875 0.011129748768688 0.01790497897082 0.015149990975236 0.017247375516567 0.014414635767837 0.018715944750698 0.040535187580391 0.041278697681637 0.041851215910992 0.042571565379811 0.036077622238471 0.041012088595584 0.041613798260117 0.041730817848975 -1.272956812892684 0.030397275562318 0.030519081096609 0.038513004999153 0.041057404969689 0.039562186859485 0.036086823397469 0.027568902482651 0.029040875908033 0.024047961286321 0.019614373795552 0.031185542975114 0.032569583432635 0.032149330115116 0.033487234799866 0.019124516914004 0.023655471435995 0.027619871002173 0.027392833475405 0.02793685726598 0.024412912074176 0.025250478251116 0.019974850031153 0.031822070815442 0.031104431259146 0.026245290126437 0.034033344331628 0.035041013454039 0.026695052472784 0.01838716404483 0.020149783475946 0.021617820007933 0.023502478080541 0.020435091737942 0.005217356554533 -0.005064876714163 -0.004301575947511 -0.003745612271114 -0.005357287737861 -0.006333662676065 -0.018296475908927 -0.032234313289406 -0.030547935546408 -0.019962117806281 -0.021706538124392 -0.021287193279028 -0.015991676139112 -0.015200094292416 -0.018858697921585 -0.00874082709014 -0.000652614726259 0.002367787018374 0.001103297091311 -0.014902287760829 -0.00401745670832 -0.013484806651829 -0.008418350645111 -0.006163251383365 -0.007076200227641 -0.001743589950168 0.012931524965025 0.007251130273739 0.013507481621106 0.00962357478905 0.007421580860819 0.015075744607183 0.013409564053238 0.012267212913084 0.007857971356325 0.006486440663547 0.011483431982981 0.005104112427025 0.005571207521907 0.010213788528161 -0.000506428875482 -0.001607733007837 -0.005838882911521 0.002340152276084 0.000262467661829 0.000465772389123 -0.007206346607822 -0.007400576800226 -0.004721776570725 -0.01100740611866 -0.007213139215336 -0.019417897546685 -0.021478632516168 -0.034938320592804 -0.036340724352632 -0.024553789201827 -0.021524093688043 -0.034430861412601 -0.03565048905559 -0.020122583810327 -0.003768948811296 -0.002932275264582 0.003196780816291 -0.001418005433942 -0.002789638818741 -0.00757472007642 -0.00835611607228 -0.018494911639164 -0.01681246580583 -0.007361846763214 -0.010376877031095 0.002726655719352 -0.004645185965992 -0.004549572873096 0.002076951056135 0.001711298796336 0.001339997909541 0.000964690617788 0.003917465857606 0.000208581786161 0.008109614693887 0.005837004864306 0.016805989492614 0.013281961327061 0.016081574390781 0.02621666531033 0.021799937526293 0.018957514525618 0.028043286146499 0.02475041037817 0.027440089208916 0.018686832529642 0.014787548447368 0.018698021120133 0.015834069201202 0.016935897458079 0.015348809558687 0.009187933114912 0.018878673994216 0.018694022460737 0.013951039262491 0.023365788737847 0.021666531591869 0.00451941534131 0.000299852160365 0.004235554443437 0.004170473249323 0.004035937578563 0.003366686221851 0.003767064198409 -0.025406671762813 -0.025547539777563 -0.02569339496697 -0.025845325877917 -0.026004242630933 -0.026639223387535 -0.024236341097744 -0.031213458038603 -0.02076253900492 -0.020288567665848 -0.019732557382747 -0.021619861175106 -0.029479149490876 0.002233659244875 -0.005625967234217 0.001748211799812 0.015423697189523 0.012864472359985 0.017061423837903 0.016213071027983 0.001956972721874 -0.003606283343673 0.011806526788712 0.010472986104458 0.005106060306237 0.004243249181111 0.005098827538648 0.018686454335946 0.013268930379905 0.012999024821448 0.013790425426088 0.020187604756171 0.022013387601819 0.014098586816214 0.005346370262958 -0.000207649963679 0.0002147357095 -0.015996961547134 -0.017303067755629 -0.016575544310866 -0.033105584277902 -0.031010629580082 -0.029994818252899 -0.016438528983476 -0.016758574085933 -0.00540282488609 -0.017413328605258 -0.023353694493011 -0.00923678319911 -0.021958674373329 -0.00695850239458 0.007441272375377 0.012887990217608 0.020934415316398 0.018876659964098 0.015213521947251 0.013207059311346 0.018844400026125 0.008938713253937 0.008600127748443 0.007210131982773 0.002897364098301 0.004270374202975 0.013430426171467 0.009009138420652 0.012877019568529 0.017089109801853 0.012602695024424 0.006866355324746 0.012102727671276 0.016753822540852 0.005907357636247 0.0006622078437 0.003020746633995 -0.009679118512926 -0.01173438194554 -0.003544760075174 -0.004437567711562 -0.008295263317579 -0.00587608563227 -0.009257592995082 -0.008115588438795 -0.009678153711785 -0.018148474089653 请问这是咋回事 怎么根据变换输出的Gam窗口重构原信号啊，关键这个我不会
好像需要把GAM扩展成M*N的矩阵
难道要对GAM进行周期延陀？ 本帖最后由 wdhd 于 2016-9-8 14:17 编辑
贴一个短时傅立叶逆变换的程序，楼主看看是否有用。
function =invspecgram(B,NFFT,Fs,WINDOW,NOVERLAP);
% INVSPECGRAM Performs the Inverse Short Time FFT (Inverse SPECGRAM).
%% Usage: = INVSPECGRAM(B,NFFT,Fs,WINDOW,NOVERLAP);
% % B is the STFT data; NFFT is the number of points for which the FFT was
% calculated per slice.
% Fs is the sampling frequency of the original data. WINDOW is the window
% length in samples; NOVERLAP is the number of samples which each FFT slide
% overlap with each other.
%% See also SPECGRAM, FFT and IFFT.
stepsize = WINDOW - NOVERLAP;
=size(B);
transB=zeros(size(b,a));
ispecgram = zeros((((stepsize*b)+a)),1);
B = ifft(B,NFFT);
transB = B';
counter3 = 1;
for counter2= 1:1:b
ispecgram(counter3:(((floor(a/2))-1)+counter3),1) = (transB(counter2,1:(floor(a/2)))') + ispecgram(counter3:(((floor(a/2))-1)+counter3),1);
counter3 = counter3 + stepsize;endNX =(b*(WINDOW-NOVERLAP))+NOVERLAP;
data =real(ispecgram) .* 2;data = data(1:NX);
[ 本帖最后由 zhlong 于 2007-11-30 19:45 编辑 ]
本帖最后由 wdhd 于 2016-9-8 14:17 编辑
原帖由 zhlong 于 2007-11-30 19:43 发表
贴一个短时傅立叶逆变换的程序，楼主看看是否有用。
这个程序怎么输入Gabor变换的高斯窗口，windows只是窗口长度啊
回复 #6 zhlong 的帖子
这个程序的运行结果非常糟糕不明白大家都是怎么做Gabor滤波的重构的 这种重构方式也不理想，根本没法用
yh是tfrgabor生成的高斯窗
yh = padarray(yh,,'circular','pre');
for k=1:length(y)
a=0;
for i=1:m
for j=1:n
ss=(k-i*length(y)/m);
if ss<=0
ss=-ss+1;
end
a=ydgr(i,j)*yh(ss)*exp(sqrt(-1)*2*pi*j*k/n)+a;
end
end
y1(k)=a;
end
页:
[1]