• No results found

Skript reliefman.m

N/A
N/A
Info
Download
Protected

Academic year: 2022

Share "Skript reliefman.m"

Copied!
10
0
0

Loading.... (view fulltext now)

Download now ( 10 Page )

Full text

(1)

Skript reliefman.m

clc;clear all;pobr=1;format compact;close all

warning off all;warning('off','MATLAB:dispatcher:InexactMatch');

clear,clc;

pocet_obrazku=150,%zadej pocet obrazku pocet_harm=30;pobr=1;

for k=118:118+pocet_obrazku name=num2str(k);

obr=imread(['obr_' num2str(k) '.tif']);

obr=obr(400:end-300,:,:);

figure,imshow(obr,[]) Gx=obr;

obr=rgb2gray(obr);

bw=histeq(obr,256);

figure,imshow(bw,[]) level2=graythresh(bw);

BW = im2bw(bw, level2);

BW_1=abs(BW-1);

figure,imshow(BW1_1);

BW1= imfill(BW1_1);

figure,imshow(BW1,[]) se=strel('disk',1);

BW3=abs(BW1-1);

for i=1:5

BW3=imerode(BW3,se);

end

figure,imshow(BW3) se=strel('disk',3);

BW4=imdilate(BW3,se);

figure,imshow(BW4,[]) BW4=abs(1-BW4);

figure,imshow(BW4,[]) BW8=BW4';

%figure,imshow(BW8) data=[];

for kk=1:size(BW8,1)

[r s]=find(BW8(kk,:)==1,1,'first');

data=[data;kk,s];

end

str_data=round(mean(data(:,2)));

%data(:,2)=data(:,2)-str_data;

data_fft=fft(data(:,2));

data_upr=data_fft;

data_upr(pocet_harm:length(data_upr)-pocet_harm,1)=0; %eliminace amplitud na vyssich frekv z obou stran spektra- suda fce

%data_upr(pocet_harm:length(data_upr)- pocet_harm,2)=0;

data_ifft_u=ifft(data_upr);

%figure,plot(data(:,1),data(:,2),'b',data(:,1),abs(data_ifft_u(:,1)),'r %');

%axis equal; %abs-je treba zobrazovat abs hodnotu re cisla

for jj=1:length(data)

Gx (data(jj,2),data(jj,1),1)=255; Gx(str_data,data(jj,1),1)=0;

(2)

Gx(round(abs(data_ifft_u(jj,1))),data(jj,1),1)=0;

Gx (data(jj,2),data(jj,1),2)=0; Gx(str_data,data(jj,1),2)=255;

Gx(round(abs(data_ifft_u(jj,1))),data(jj,1),2)=0;

Gx (data(jj,2),data(jj,1),3)=0; Gx(str_data,data(jj,1),3)=0;

Gx(round(abs(data_ifft_u(jj,1))),data(jj,1),3)=255;

end

figure,imshow(Gx,[]) end

(3)

Skript drsnostRCM.m

clc;clear all;pobr=1;format compact;close all

warning off all;warning('off','MATLAB:dispatcher:InexactMatch');

clear,clc;

pocet_obrazku=150; %zadej pocet obrazku pocet_harm=30;pobr=1;

sloupec=3;%-- 8.12.11 12:16 --%

cit=1;

vysledky=[];

data_matice=[];

data_matice_str=[];

for k=100:100+pocet_obrazku name=num2str(k);

obr=imread(['a1t' num2str(k) '.tif']);

%obr=obr(400:end-300,:,:);

%figure,imshow(obr,[]) Gx=obr;

obr=rgb2gray(obr);

bw=histeq(obr,256);

%figure,imshow(bw,[]) level2=graythresh(bw);

BW = im2bw(bw,level2);

BW1_1=abs(BW-1);

%figure,imshow(BW1_1,[]) BW1 = imfill(BW1_1);

%figure,imshow(BW1,[]) se = strel('disk',1);

BW3=abs(BW1-1);

for i=1:5

BW3 = imerode(BW3,se);

end

%figure,imshow(BW3) se = strel('disk',3);

BW4 = imdilate(BW3,se);

%figure,imshow(BW4,[]) BW4=abs(1-BW4);

%figure,imshow(BW4,[]) BW8=BW4';

%BW8=abs(1-BW8);

%figure,imshow(BW8) data=[];

for kk=1:size(BW8,1)

[r s]=find(BW8(kk,:)==1,1,'first');

data=[data; kk s];

end

data(:,2)=-1*data(:,2);

data_fft=fft(data(:,2));

data_upr=data_fft;

data_upr(pocet_harm:length(data_upr)-pocet_harm,1)=0; %eliminace amplitud na vyssich frekv z obou stran spektra - suda fce

data(:,3)=-1*abs(ifft(data_upr));

str_data=(mean(data(:,3)));

data(:,3)=data(:,sloupec)-str_data;

yna=detrend(data(:,3));

yva=detrend(abs(data(:,3)));

pyva=detrend(data(:,2));

man(k,:)=yna;

mav(k,:)=yva;

pov(k,:)=pyva;

(4)

yna1=detrend(data(:,2));

%yva1=detrend(abs(data(:,2)));man1(k,:)=yna1;mav1(k,:)=yva1;

%yna=max(yna)-yna/(max(yna)-min(yna));yva=max(yva)-yva/(max(yva)- min(yva));%mass(k,:)=yva;

str_data=(mean(data(:,sloupec)));

data_matice=[data_matice; k data(:,sloupec)'];

%data_matice_str=[data_matice; k (data(:,sloupec)-str_data)'];

data(:,2)=data(:,2)-mean(data(:,2));

data(:,sloupec)=data(:,sloupec)-str_data;

str_data=(mean(data(:,sloupec)));

data(:,4)=data(:,sloupec)*-1+2*str_data;

data(:,5)=data(:,sloupec)>str_data;

data(:,5)=data(:,5).*data(:,sloupec);

data(:,6)=data(:,sloupec+1)>str_data;

data(:,6)=data(:,6).*data(:,sloupec+1);

MX=[ones(length(data),1)];

scp=(inv(MX'*MX)*MX'*data(:,sloupec));

roz=(data(:,sloupec)-scp);

rozsort=sort(roz,'descend');

MAD=1/length(data)*(sum(abs(roz)));

Rmax=max(roz); [ind1]=find(roz==Rmax);

Rmin=min(roz); [ind2]=find(roz==Rmin);

Rm=Rmax-Rmin;

TP=(sum(rozsort(1:5))+sum(rozsort(end-4:end)))/10;

[ind3]=find(round(data(:,sloupec))==round(scp));

pom=diff(ind3);

Zm=mean(pom(pom~=1));

[ind4 c1]=findpeaks(data(:,5));

Z=mean(diff(c1));

pomtp=[];

for kk=1:length(data)-1

pomtp=[pomtp; kk pdist([data(kk:kk+1,1) data(kk:kk+1,3)])];

end

tp=sum(pomtp(:,2))/length(data);

P=ind4-scp; % [ind5]=find(P<0); P(ind5)=0;

MP=mean(P);

[ind6 c2]=findpeaks(data(:,6));

V=ind6-scp; %[ind7]=find(V<0); V(ind7)=0;

MV=mean(V);

SD=std(data(:,sloupec));

CV=SD/MAD;

PSC=sqrt(mean((diff(data(:,3)).^2)));

PC=sqrt(mean((diff(data(:,3),2).^2)));

MS=mean(abs(diff(data(:,3))));

vysledky=[vysledky; cit str_data scp MAD Rm TP Zm Z tp MP MV SD CV MS PSC PC MS];

cit=cit+1

%vektor=data(:,2)';

%vysl=[];

%n=10;

%for i=1+n:length(data)-n

%if vektor(i)>vektor(i-n:i-1) & vektor(i)>vektor(i+1:i+n)

(5)

%vysl=[vysl; i vektor(i)];

%end %end %vysl

% for jj=1:length(data)

% Gx(round(data(jj,2)),round(data(jj,1)),1)=255;

Gx(str_data,data(jj,1),1)=0;

%Gx(round(abs(data_ifft_u(jj,1))),data(jj,1),1)=0;

% Gx(round(data(jj,2)),round(data(jj,1),2))=0;

Gx(str_data,data(jj,1),2)=255;

%Gx(round(abs(data_ifft_u(jj,1))),data(jj,1),2)=0;

% Gx(round(data(jj,2)),round(data(jj,1)),3)=0;

Gx(str_data,data(jj,1),3)=0;

%Gx(round(abs(data_ifft_u(jj,1))),data(jj,1),3)=255;

% end

%figure,imshow(Gx,[])

% %figure

%plot(data(:,1),data(:,2),'b'),hold on %plot(data(:,1),data(:,3),'r')

%plot(data(:,1),str_data*ones(length(data),1),'k') %plot([ind1(1) ind1(1)],[scp scp+Rmax],'g')

%plot([ind2(1) ind2(1)],[scp scp+Rmin],'g') end

%man(1:100,:)=[];

%figure

%surf(man);colorbar

%shading interp

%colormap jet

%view(-13,72)

%pov(1:100,:)=[];

%figure

%surf(pov);colorbar

%shading interp

%colormap jet

%view(-13,72)

%figure %mesh(mav);

%colorbar;

%figure %surfc(mav);

%mass=(mav);

%contour(mass)

%

%figure %mesh(mav1);

%colorbar;

%figure

%surfc(mav1);

%mass1=(mav1);

%contour(mass1)

%mav (1:100,:)=[];

%figure

(6)

%mesh(max(mav(:))-mav);

%colorbar;

%figure

%surfc(max(mav(:))-mav);

%mass=(max(mav(:))-mav);

%contour(mass);colormap(jet);colorbar;c1=1;title('height variation');

%figure

%[X Y]=meshgrid(1:1268,1:101);

%Z=data_matice_str(2:end,2:end);

%surf(X,Y,Z) %shading interp

(7)

Skript drsnostT.m

clear,clc,close all load vzorka_A.mat pocet_harm=30;

sloupec=3;

A=M(:,3);A(end)=[];

B = reshape(A,1501,1000);

B=B';

[r s]=find(B<3000);

for i=1:length(r) B(r(i),s(i))=NaN;

end

pom=isnan(B);

[r s]=find(pom==1);

ok=5;

for i=1:length(r)

if r(i)>size(B,1)-ok

B(r(i),s(i))=nanmean(nanmean(B(r(i)-ok:r(i),s(i)-ok:s(i))));

elseif s(i)>size(B,2)-ok

B(r(i),s(i))=nanmean(nanmean(B(r(i)-ok:r(i),s(i)-ok:s(i))));

else B(r(i),s(i))=nanmean(nanmean(B(r(i):r(i)+ok,s(i):s(i)+ok)));

end end

%figure,imshow(B,[]) %C=mat2gray(B);

%C=round(C*100)/100;

%figure,imshow(C,[]) cit=1;

vysledky=[];

data_matice=[];

data_matice_str=[];

data(:,1)=1:size(B,2);

for k=1:size(B,1)

data_fft=fft(B(k,:));

data_upr=data_fft;

data_upr(1,pocet_harm:length(data_upr)-pocet_harm)=0; %eliminace amplitud na vyssich frekv z obou stran spektra - suda fce

data(:,2)=B(k,:)';

str_data=(mean(data(:,2)));

data(:,2)=data(:,2)-str_data;

data(:,3)=abs(ifft(data_upr))';

str_data=(mean(data(:,3)));

data(:,3)=data(:,3)-str_data;

yna=detrend(data(:,3));

yva=detrend(abs(data(:,3)));

man(k,:)=yna;

mav(k,:)=yva;

yna=max(yna)-yna/(max(yna)-min(yna));

yva=max(yva)-yva/(max(yva)-min(yva));%mass(k,:)=yva;

str_data=(mean(data(:,sloupec)));

data_matice=[data_matice; k data(:,sloupec)'];

data_matice_str=[data_matice; k (data(:,sloupec)-str_data)'];

data(:,4)=data(:,sloupec)*-1+2*str_data;

data(:,5)=data(:,sloupec)>str_data;

data(:,5)=data(:,5).*data(:,sloupec);

(8)

data(:,6)=data(:,sloupec+1)>str_data;

data(:,6)=data(:,6).*data(:,sloupec+1);

MX=[ones(length(data),1)];

scp=(inv(MX'*MX)*MX'*data(:,sloupec));

roz=(data(:,sloupec)-scp);

rozsort=sort(roz,'descend');

MAD=1/length(data)*(sum(abs(roz)));

Rmax=max(roz); [ind1]=find(roz==Rmax);

Rmin=min(roz); [ind2]=find(roz==Rmin);

Rm=Rmax-Rmin;

TP=(sum(rozsort(1:5))+sum(rozsort(end-4:end)))/10;

pom=[];

for ii=1:size(data,1)-1 if data(ii,3)*data(ii+1,3)<0 pom=[pom; ii];

end end

%[ind3]=find(round(data(:,sloupec))==round(scp));

pom=diff(pom);

Zm=mean(pom(pom~=1));

[ind4 c1]=findpeaks(data(:,5));

Z=mean(diff(c1));

pomtp=[];

for kk=1:length(data)-1

pomtp=[pomtp; kk pdist([data(kk:kk+1,1) data(kk:kk+1,3)])];

end

tp=sum(pomtp(:,2))/length(data);

P=ind4-scp; % [ind5]=find(P<0); P(ind5)=0;

MP=mean(P);

[ind6 c2]=findpeaks(data(:,6));

V=ind6-scp; %[ind7]=find(V<0); V(ind7)=0;

MV=mean(V);

SD=std(data(:,sloupec));

CV=SD/MAD;

PSC=sqrt(mean((diff(data(:,3)).^2)));

PC=sqrt(mean((diff(data(:,3),2).^2)));

MS=mean(abs(diff(data(:,3))));

vysledky=[vysledky; cit str_data scp MAD Rm TP Zm Z tp MP MV SD CV MS PSC PC MS];

cit=cit+1

%figure

%plot(data(:,1),data(:,2),'b'),hold on %plot(data(:,1),data(:,3),'r')

%plot(data(:,1),str_data*ones(length(data),1),'k') %plot([ind1(1) ind1(1)],[scp scp+Rmax],'g')

%plot([ind2(1) ind2(1)],[scp scp+Rmin],'g') end

[X Y]=meshgrid(0.1:0.1:150.1,0.1:0.1:100);

figure,surf(X,Y,B) shading interp colormap jet

(9)

colorbar view(-13,72)

figure

surf(man);colorbar shading interp colormap jet view(-13,72)

mass=(max(mav(:))-mav);

contour(mass);colormap(jet);colorbar;c1=1;title('height variation');

(10)

Skript vyhodnoceni.m

clear,clc,close all load vysledkyRCMa.mat

% kalibrace 1px=... mm

kal=(1263/(2430/25.4))/1263;

kal2=0.03;

RCM=vysledky*kal2;

RCM(:,1:3)=[];

mRCM=mean(RCM);

mRCM([2 4 5])=[];

load vysledkyTa.mat stupen=1;

char=strvcat(' ','x','x^2','x^3','x^4');

T=vysledky*kal;

T(:,1:3)=[];

aaa=isnan(T);

[rr ss]=find(aaa==1);

T(rr,:)=[];

mT=nanmean(T);

mT([2 4 5])=[];

clear vysledky

plot(mRCM,mT,'o'),hold on corrcoef(mRCM,mT)

X=[ones(length(mT),1) mRCM'];

[b,bint,r,rint,stats] = regress(mT',X);

xx=min(mRCM):0.01:max(mRCM);

yy=polyval(flipud(b),xx);

plot(xx,yy,'r') rovnice=[];

for i=1:stupen+1

if b(i)>0; a='+ '; else a='- '; end

rovnice=[rovnice,a num2str(abs(b(i))) char(i,:)];

end

R2=sqrt(stats(1));

text(0.05,6,['y = ' rovnice],'FontSize',10) % 0.05,1.5 text(0.05,5.5,['R = ' num2str(R2)],'FontSize',10) % 0.05,1.3

[muhat1,sigmahat1,muci1,sigmaci1] = normfit(RCM,0.05);

[muhat2,sigmahat2,muci2,sigmaci2] = normfit(T,0.05);

figure,boxplot(RCM,'notch','on'), figure,boxplot(T,'notch','on'), e1=muci1(2,:)-muhat1;

e2=muci2(2,:)-muhat2;

figure,errorbar(1:14,muhat1,e1,'x'),hold on errorbar(1:14,muhat2,e2,'rx')

%vys=[];

%for j=1:size(RCM,2)

%[h,p,ci,stats] = ttest2(RCM(:,j),T(:,j),0.05) %jbtest(RCM(:,j),0.05);

%vys=[vys; h p ci' stats.tstat stats.df stats.sd];

%end %vys

References

Download now ( PDF - 10 Page - 77.35 KB )

Related documents

“Close Encounters” Spelkoncept:

Nästan allt av det man kan hitta i världen kommer du att kunna interegera med, till exempel så kommer du att kunna använda stenar och pinnar som temporära vapen om du inte har

Mat

Må- tyås Szabö, själv invandrare, tecknar bilden av deras bidrag till vårt kosthåll idag. Och visst ger deras butiker även oss infödda svenskar en möjlighet att få tag på

Violence in close relationships

Crisis Centre in Central Skåne(=Kriscentrum Mellersta Skåne) 046-35 50 71 If you live in Höör, you can turn to Crisis Centre in Central Skåne, a receiving for dialogue that

Vad blir det för mat?

Några av intervjupersonerna berättar även att de emellanåt lagar maten själva på helgerna och upplever då att de boende äter mer, vilket i så fall kanske kan påverka

”Vad blir det för mat?”

Ungdomar har en positiv attityd till hem- och konsumentkunskap då de där ges tillfälle att öva sina färdigheter och får möjlighet att vara delaktiga vilket skulle kunna ge dem

Jak se provádí kalibrace tisku? Student odpověděl

Jaké jsou reálné rychlosti tisku a jaký objem za jednotku času jste schopen tisknout.. Student

Povrchové šet ení koncových m rek p i interferen ní metod kalibrace v MI OI Liberec

Dále byla pro povrchové šet ení použita metoda mikroskopie povrchu AFM (Atomic Force Microscopy – Mikroskopie atomových sil) využívající slabých meziatomových sil. V

Clear Round

The gendering of different stable environments is then seemingly racialized in specific ways, making English riding disciplines (especially dressage) habitually appear as white