Sie sind auf Seite 1von 58

Fundamental of Image Processing

8th Sem - E & C

Practical No: 1 Aim: Introduction to image processing commands. Program:


clc; clear all; close all; %% %to read grey scale image I = imread('pout.tif'); % imshow(I); %% % to read colour image RGB = imread('football.jpg'); %imshow(RGB); subplot(1,2,1);imshow(I);title('gray'); subplot(1,2,2);imshow(RGB);title('colour'); %% %to convert colour to grey GR = rgb2gray(RGB); % figure, imshow(GR); %to convert grey to binary BR = im2bw(GR); % figure, imshow(BR); %to complement the colour image CPL = imcomplement(RGB); CPL1 = imcomplement(GR); CPL2 = imcomplement(BR); % imshow(CPL); figure; subplot(2,3,1);imshow(RGB);title('colour'); subplot(2,3,2);imshow(GR);title('gray'); subplot(2,3,3);imshow(BR);title('binary'); subplot(2,3,4);imshow(CPL);title('negative of colour'); subplot(2,3,5);imshow(CPL1);title('negative of gray'); subplot(2,3,6);imshow(CPL2);title('negative of binary'); %% %to rotate the image RT1 = imrotate(RGB,+45); RT2 = imrotate(RGB,-45); figure; subplot(1,3,1);imshow(RGB);title('colour'); subplot(1,3,2);imshow(RT1);title('rotation45degree'); subplot(1,3,3);imshow(RT2);title('rotation-45degree'); %% %to know image information IF =imfinfo('football.png');

Date: 11/02/2013

CGPIT

Department of Electronics & Communication Engineering

Page 1

Fundamental of Image Processing

8th Sem - E & C

Result & Conclusion:


Figure: 1: Gray & Color Image.

Figure: 2: Negative of different Images.

CGPIT

Department of Electronics & Communication Engineering

Page 2

Fundamental of Image Processing Figure: 3: Rotation of Image.

8th Sem - E & C

Work Space:

CGPIT

Department of Electronics & Communication Engineering

Page 3

Fundamental of Image Processing Image Information using imfinfo:

8th Sem - E & C

Conclusion: From this Experiment we have concluded that we can change parameters of any
images with only one simple function using MATLAB

CGPIT

Department of Electronics & Communication Engineering

Page 4

Fundamental of Image Processing

8th Sem - E & C

Practical No: 2 Aim:


A. B. C. D. To read color image and convert it into gray scale. To read gray scale image and convert it into binary scale. Convert binary image into negative image. Apply spatial resolution and gray level resolution to gray scale image.

Date: 18/02/2013

Program:
clc; clear all; close all; t=imread('peppers.png'); subplot(1,3,1);imshow(t);title('colour'); r=t(:,:,1); g=t(:,:,2); b=t(:,:,3); Y=0.2989 * r + 0.5870 * g + 0.1140 * b; subplot(1,3,2);imshow(Y);title('gray'); [m,n]=size(Y); z=zeros(m,n); for i=1:m for j=1:n if Y(i,j)<128 z(i,j)=0; else z(i,j)=1; end end end subplot(1,3,3);imshow(z);title('binary'); %%

%to convert gray to negative L = 255-Y; figure,subplot(1,2,1);imshow(L);title('negative of colour');

%binary to negative G = 1-z; subplot(1,2,2);imshow(G);title('negative of binary');

CGPIT

Department of Electronics & Communication Engineering

Page 5

Fundamental of Image Processing %%spatial resolution B1 = imresize(t,[192,256]); B2 = imresize(t,[90,128]); B3 = imresize(t,[45,64]); figure; subplot(2,2,1);imshow(t); subplot(2,2,2);imshow(B1); subplot(2,2,3);imshow(B2); subplot(2,2,4);imshow(B3);

8th Sem - E & C

%%intensity resolution figure; t1=imread('C:\Program Files\MATLAB\R2010a\toolbox\images\imdemos\html\ctskull.tif'); for i=1:8 J=t1/(2^i); subplot(2,4,i);imshow(J,[]); end

Result & Conclusion:


Figure: 1: Display different types of image.

CGPIT

Department of Electronics & Communication Engineering

Page 6

Fundamental of Image Processing Figure: 2: Negative of Colour & Binary Image.

8th Sem - E & C

Figure: 3: Spatial Resolution.

CGPIT

Department of Electronics & Communication Engineering

Page 7

Fundamental of Image Processing Figure: 4: Gray Level Intensity Resolution.

8th Sem - E & C

Conclusion:
In part(A), we have converted the Original Color image into Graylevel image by adding the different planes of original color image. In part (B), we have converted Graylevel image into binary image by converting 8 bit image into 2-bit image. In part (C), we have complemented the binary image to get the negative of binary image. In part(D), Spatial Resolution is done by resizing the Original image, whereas Intensity Resolution represents the image having decreased Intensity levels.

CGPIT

Department of Electronics & Communication Engineering

Page 8

Fundamental of Image Processing

8th Sem - E & C

Practical No: 3 Aim:

Date: 25/02/2013

A. Intensity transformation using Negative, Log and Gamma (power law) transformation of Gray Scale Image. Comment on results. B. Take 4 types of images: dark, light, high contrast, low contrast and compare their histogram. Comment on results. C. Enhance contrast of an intensity image using histogram equalization of Part (b). Display and compare the histogram of the original and processed image.

Program:
clc; clear all; close all;

%% To find Negative Image I=imread('tire.tif'); figure(); subplot(121); imshow(I); title('original image'); J=imcomplement(I); subplot(122); imshow(J); title('negative image');

%% To find Histogram of an Image

a=imread('tire.tif'); A=double(a); [m n]=size(A); L=zeros(256,1); for i=1:m; for j=1:n; for k=1:256; if (A(i,j)==k) L(k)=L(k)+1; end end end CGPIT

%% it can store large number of variable %% for finding number of rows and columns

%% for 256 intensity levels

Department of Electronics & Communication Engineering

Page 9

Fundamental of Image Processing end

8th Sem - E & C

subplot(121); stem(L); %% for ploting the histogram title('calculated histrogram'); subplot(122); imhist(a); %% for ploting the histogram title('direct histrogram');

%% To find Log Transform I=imread('tire.tif'); I1=im2double(I); M1=1*log(1+I1); M2=2*log(1+I1); M3=5*log(1+I1); figure(); subplot(221); imshow(I); title('Original Image'); subplot(222); imshow(M1) title('Image with c=1'); subplot(223); imshow(M2) title('Image with c=2'); subplot(224); imshow(M3) title('Image with c=3');

%% s=c*log(1+r)

%% To Perform Gamma Transformation I=imread('tire.tif'); J1=imadjust(I,[],[],1); CGPIT %% s=c*r^gamma


Page 10

Department of Electronics & Communication Engineering

Fundamental of Image Processing J2=imadjust(I,[],[],3); J3=imadjust(I,[],[],0.4); figure(); subplot(131); imshow(J1); title('image with gamma =1'); subplot(132); imshow(J2); title('image with gamma =3'); subplot(133); imshow(J3); title('image with gamma =0.4');

8th Sem - E & C

%% To Perform Histogram Equilization I = imread('tire.tif'); figure,imshow(I); J = histeq(I); %% Histogram Equalization subplot(2,2,1); imshow(I); title('Original Image'); subplot(2,2,2); imshow(J); title('Processed Image');

%% Processed image

subplot(2,2,3); imhist(I,64); title('Histogram of Original Image'); %% Histogram of the original image. subplot(2,2,4); imhist(J,64); title('Histogram of Processed Image'); %% Histogram of the processed image

CGPIT

Department of Electronics & Communication Engineering

Page 11

Fundamental of Image Processing

8th Sem - E & C

Result & Conclusion:


Figure: 1: Histogram

Figure: 2: Log Transformation.

CGPIT

Department of Electronics & Communication Engineering

Page 12

Fundamental of Image Processing Figure: 3: Gamma Transformation.

8th Sem - E & C

Figure: 4: Enhance contrast of an intensity of image using Histogram Equalization.

CGPIT

Department of Electronics & Communication Engineering

Page 13

Fundamental of Image Processing Part (B) & (C) histogram and equalized histogram 4 different images

8th Sem - E & C

Conclusion: In part (A) we have converted Gray level image into negative image by
complementing the intensity values. When we apply Log Transform to the original image, darker portion of image is scaled in brighter portion. In Gamma Transformation, if gamma<1, it will gives the brighter image while gamma>1 gives darker image than original one. In part (B)we have read the 4- types of images and plotted the histogram and we can see that intensity levels are in random nature. In part (C) we have performed the Histogram equalization of image and observe that intensity levels are well distributed overall

CGPIT

Department of Electronics & Communication Engineering

Page 14

Fundamental of Image Processing

8th Sem - E & C

Practical No: 4

Date: 04/03/2013

Aim: To perform bit plane slicing of an 8 bit image and reconstruct it again. Program:
clc; close all; clear all; I=imread('C:\Program Files\MATLAB\R2010a\toolbox\images\imdemos\cameraman.tif'); imshow(I); [row,col]=size(I); bit_slicing=zeros(row,col); % bit_slicing(:,:)=255; % n=input('enter the value of bit:');

figure, for n = 1:8 for x=1:1:row for y=1:1:col bit_slicing(x,y,n)= bitand(I(x,y),2^(n)); end end subplot(3,3,n); imshow(uint8(bit_slicing(:,:,n)),[]); end %Image reconstruction by combining 8 bit planes A=imread(':\Program Files\MATLAB\R2010a\toolbox\images\imdemos\cameraman.tif'); B=zeros(size(A)); B=bitset(B,8,bitget(A,8)); B=bitset(B,7,bitget(A,7)); B=bitset(B,6,bitget(A,6)); B=bitset(B,5,bitget(A,5)); B=bitset(B,5,bitget(A,4)); B=bitset(B,5,bitget(A,3)); B=bitset(B,5,bitget(A,2)); B=bitset(B,5,bitget(A,1)); B=uint8(B); figure,imshow(B);
Page 15

CGPIT

Department of Electronics & Communication Engineering

Fundamental of Image Processing

8th Sem - E & C

Result & Conclusion:


Figure: 1: Original 8 bit image.

Figure: 2: Reconstructed 8 bit image.

CGPIT

Department of Electronics & Communication Engineering

Page 16

Fundamental of Image Processing Figure: 3: Bit Plane Slicing.

8th Sem - E & C

Conclusion: From this Experiment we have done bit plane slicing of 8 bit Gray Scale image &
reconstruct it from these 8 bit slices of 8 bit images using bitset & bitget functions.

CGPIT

Department of Electronics & Communication Engineering

Page 17

Fundamental of Image Processing

8th Sem - E & C

Practical No: 5 Aim: To Implement Smoothing spatial filter

Date: 11/03/2013

- Read an image and then corrupt it by salt and pepper, Gaussian, speckle noise then apply A. AVERAGING filter with 3x3 mask and 5X5 mask. B. MEDIAN filter with 3x3 and 5x5 mask. C. Max, Min, Midpoint filter with 3x3 and 5x5 mask. - Comment on the result

Program:
clc; clear all; close all;

I = imread('cameraman.tif'); B=imnoise(I,'salt & pepper'); C=imnoise(I,'gaussian'); D=imnoise(I,'speckle'); % Salt & Pepper Noise subplot(121);imshow(I),title('Original Image'); subplot(122);imshow(B),title('Salt & Pepper Noise'); h = fspecial('average',[3 3]); I2 = filter2(h,B)/255; figure();subplot(121); imshow(I2),title('Averaging Filter 3*3'); h1 = fspecial('average',[5 5]); I3 = filter2(h1,B)/255; subplot(122);imshow(I3);title('Averaging Filter 5*5'); h2 = medfilt2(B,[3 3]); h3 = medfilt2(B,[5 5]); figure,subplot(121); imshow(h2),title('Median Filter 3*3'); subplot(122);imshow(h3),title('Median Filter 5*5'); figure(); f = ordfilt2(B, 1, ones(3, 3)); f1 = ordfilt2(B, 9, ones(3, 3)); f2 = 0.5 * (f + f1); subplot(131);imshow(f);title('MIN FILTER 3*3'); subplot(132);imshow(f1);title('MAX FILTER 3*3'); subplot(133);imshow(f2);title('MIDPOINT FILTER 3*3');

CGPIT

Department of Electronics & Communication Engineering

Page 18

Fundamental of Image Processing figure(); f5 = ordfilt2(B, 1, ones(5, 5)); f15 = ordfilt2(B, 9, ones(5, 5)); f25 = 0.5 * (f5 + f15); subplot(131);imshow(f5);title('MIN FILTER 5*5'); subplot(132);imshow(f15);title('MAX FILTER 5*5'); subplot(133);imshow(f25);title('MIDPOINT FILTER 5*5');

8th Sem - E & C

% Gaussian Noise subplot(121);imshow(I),title('Original Image'); subplot(122);imshow(C),title('Gaussian Noise'); g = fspecial('average',[3 3]); I1G = filter2(g,C)/255; figure();subplot(121); imshow(I1G),title('Averaging Filter 3*3'); h1G = fspecial('average',[5 5]); I3G = filter2(h1G,C)/255; subplot(122);imshow(I3G);title('Averaging Filter 5*5'); h2G = medfilt2(C,[3 3]); h3G = medfilt2(C,[5 5]); figure,subplot(121); imshow(h2G),title('Median Filter 3*3'); subplot(122);imshow(h3G),title('Median Filter 5*5'); figure(); fG = ordfilt2(C, 1, ones(3, 3)); f1G = ordfilt2(C, 9, ones(3, 3)); f2G = 0.5 * (fG + f1G); subplot(131);imshow(fG);title('MIN FILTER 3*3'); subplot(132);imshow(f1G);title('MAX FILTER 3*3'); subplot(133);imshow(f2G);title('MIDPOINT FILTER 3*3'); figure(); f5G = ordfilt2(C, 1, ones(5, 5)); f15G = ordfilt2(C, 9, ones(5, 5)); f25G = 0.5 * (f5G + f15G); subplot(131);imshow(f5G);title('MIN FILTER 5*5'); subplot(132);imshow(f15G);title('MAX FILTER 5*5'); subplot(133);imshow(f25G);title('MIDPOINT FILTER 5*5');

CGPIT

Department of Electronics & Communication Engineering

Page 19

Fundamental of Image Processing % Speckle Noise subplot(121);imshow(I),title('Original Image'); subplot(122);imshow(D),title('Speckle Noise'); hs = fspecial('average',[3 3]); I2s = filter2(hs,D)/255; figure();subplot(121); imshow(I2s),title('Averaging Filter 3*3'); h1s = fspecial('average',[5 5]); I3s = filter2(h1s,D)/255; subplot(122);imshow(I3s);title('Averaging Filter 5*5'); h2s = medfilt2(D,[3 3]); h3s = medfilt2(D,[5 5]); figure,subplot(121); imshow(h2s),title('Median Filter 3*3'); subplot(122);imshow(h3s),title('Median Filter 5*5'); figure(); fs = ordfilt2(D, 1, ones(3, 3)); f1s = ordfilt2(D, 9, ones(3, 3)); f2s = 0.5 * (fs + f1s); subplot(131);imshow(fs);title('MIN FILTER 3*3'); subplot(132);imshow(f1s);title('MAX FILTER 3*3'); subplot(133);imshow(f2s);title('MIDPOINT FILTER 3*3'); figure(); f5s = ordfilt2(D, 1, ones(5, 5)); f15s = ordfilt2(D, 9, ones(5, 5)); f25s = 0.5 * (f15s + f5s); subplot(131);imshow(f5s);title('MIN FILTER 5*5'); subplot(132);imshow(f15s);title('MAX FILTER 5*5'); subplot(133);imshow(f25s);title('MIDPOINT FILTER 5*5');

8th Sem - E & C

CGPIT

Department of Electronics & Communication Engineering

Page 20

Fundamental of Image Processing

8th Sem - E & C

Result & Conclusion:


Salt & Pepper Noise Salt & Pepper Noise

Averaging Filter on Salt & Pepper Noise Image:

CGPIT

Department of Electronics & Communication Engineering

Page 21

Fundamental of Image Processing Median Filter on Salt & Pepper Noise Image:

8th Sem - E & C

Minimum,Maximum & Midpoint(3*3) Filter on Salt & Pepper Noise Image:

CGPIT

Department of Electronics & Communication Engineering

Page 22

Fundamental of Image Processing Minimum,Maximum & Midpoint(5*5) Filter on Salt & Pepper Noise Image:

8th Sem - E & C

Gaussian Noise Gaussian Noise:

CGPIT

Department of Electronics & Communication Engineering

Page 23

Fundamental of Image Processing Averaging Filter on Gaussian Noise Image:

8th Sem - E & C

Median Filter on Gaussian Noise Image:

CGPIT

Department of Electronics & Communication Engineering

Page 24

Fundamental of Image Processing Minimum,Maximum & Midpoint(3*3) Filter on Gaussian Noise Image:

8th Sem - E & C

Minimum,Maximum & Midpoint(5*5) Filter on Gaussian Noise Image:

CGPIT

Department of Electronics & Communication Engineering

Page 25

Fundamental of Image Processing Speckle Noise Speckle Noise:

8th Sem - E & C

Averaging Filter on Speckle Noise Image:

CGPIT

Department of Electronics & Communication Engineering

Page 26

Fundamental of Image Processing Median Filter on Speckle Noise Image:

8th Sem - E & C

Minimum,Maximum & Midpoint(3*3) Filter on Speckle Noise Image:

CGPIT

Department of Electronics & Communication Engineering

Page 27

Fundamental of Image Processing Minimum,Maximum & Midpoint(5*5) Filter on Speckle Noise Image:

8th Sem - E & C

Conclusion:
We observe that smoothing makes the image blurred. When we apply average filtering the image is more blurred and it also removes the noise. When we apply the non linear filters like median, max and min filters then image is less blurred but it is more effective on the noise compare to averaging

CGPIT

Department of Electronics & Communication Engineering

Page 28

Fundamental of Image Processing

8th Sem - E & C

Practical No: 6 Aim: To Implement Sharpening spatial filter

Date: 18/03/2013

A. To read an image and apply : Laplacian Second order filters B. To read an image and apply : Unsharp masking and High Boost filtering C. To read an image and apply : Gradient first order filters (Sobel and Prewitt Operator)

Program:
clc;clear all;close all; s=imread('cameraman.tif'); imshow(s);title('Original Image','FontSize',13); figure(); % % Laplacian Transform of Image lf1=fspecial('laplacian',0.5); lf=imfilter(s,lf1); imshow(lf);title('Image after laplacian Filter','FontSize',13); figure(); lf2=s-lf; imshow(lf2);title('Enhanced Image after laplacian','FontSize',13); figure(); % % Unsharp Masking of Image % uf1=fspecial('unsharp'); uf1=[-1 -1 -1;-1 8 -1;-1 -1 -1]; uf=imfilter(s,uf1); uf_image=s-uf; subplot(231); imshow(s); title('Original Image','FontSize',13); subplot(232); imshow(uf);title('Unsharp Masking of Image','FontSize',13); subplot(233); imshow(uf_image);title('Image after Unsharp Masking','FontSize',13); % % High Boost Filtering of Image hb1=[-1 -1 -1;-1 10 -1;-1 -1 -1]; hb=imfilter(s,hb1); hb_image=s-hb; subplot(234); imshow(s); title('Original Image','FontSize',13); subplot(235); imshow(hb_image); title('HighBoost Masking','FontSize',13); subplot(236); imshow(hb); title('Image After HighBost Filtering','FontSize',13); figure(); subplot(231); CGPIT Department of Electronics & Communication Engineering
Page 29

Fundamental of Image Processing imshow(s);title('Origanal image','FontSize',13); % %Horizontal Sobel Filter sb=fspecial('sobel'); sh=imfilter(s,sb); subplot(232); imshow(sh);title('Horizontal Sobel','FontSize',13); % % Vertical Sobel (vs=hs') Filter sb=sb'; sv=imfilter(s,sb); subplot(233); imshow(sv);title('Vertical Sobel','FontSize',13); % % Prewitt Horizontal Filter p=fspecial('prewitt'); ph=imfilter(s,p); subplot(234); imshow(ph);title('Horizontal Prewitt','FontSize',13); % % Vertical Prewitt (vp=hp') Filter p=p'; pv=imfilter(s,p); subplot(235) imshow(pv);title('Vertical Prewitt','FontSize',13);

8th Sem - E & C

Result & Conclusion: Part: A


Original Image:

CGPIT

Department of Electronics & Communication Engineering

Page 30

Fundamental of Image Processing Laplacian Filter on Image:

8th Sem - E & C

Enhanced Image after Laplacian Filter:

CGPIT

Department of Electronics & Communication Engineering

Page 31

Fundamental of Image Processing

8th Sem - E & C

Part: B

CGPIT

Department of Electronics & Communication Engineering

Page 32

Fundamental of Image Processing

8th Sem - E & C

Part: C

Conclusion:
After completion of this practical, we analyze that sharpening enhances the sweep changes in the intensity in the image and such changes becomes clearer in image. So, here details like edges of objects and lines enhanced

CGPIT

Department of Electronics & Communication Engineering

Page 33

Fundamental of Image Processing

8th Sem - E & C

Practical No: 7 Aim: To implement filters in frequency domain

Date: 25/03/2013

A. To read an image and apply: Low pass Ideal, Butterworth and Gaussian filter B. To read an image and apply: High pass Ideal, Butterworth and Gaussian filter Use cut-off frequency 5,10,15,80 and 230 for all above filters.

Program:
clc;clear all;close all; i=imread('cameraman.tif'); i1=fftshift(fft2(i)); s=input('Enter Order of Filter: '); [m n]=size(i1); D=0;H_5=0;H_15=0;H_30=0;H_80=0;H_230=0; for u=1:m for v=1:n; D(u,v)=sqrt((u-(m/2))^2+(v-(n/2))^2); % %Low pass Butterworth Filter LHB_5(u,v)=(1/(1+(D(u,v)/5)^2*s)); LHB_50(u,v)=(1/(1+(D(u,v)/50)^2*s)); LHB_200(u,v)=(1/(1+(D(u,v)/200)^2*s)); % %Low pass Gausian Filter LHG_5(u,v)=exp(-(D(u,v)^2/(2*(5)^2))); LHG_50(u,v)=exp(-(D(u,v)^2/(2*(50)^2))); LHG_200(u,v)=exp(-(D(u,v)^2/(2*(200)^2))); % %Ideal Low pass Filter if D(u,v)<=5; LH_5(u,v)=1; else LH_5(u,v)=0; end if D(u,v)<=50; LH_50(u,v)=1; else LH_50(u,v)=0; end if D(u,v)<=200; LH_200(u,v)=1; else LH_200(u,v)=0; end end end CGPIT Department of Electronics & Communication Engineering
Page 34

Fundamental of Image Processing % %Ideal Low pass Filter subplot(321); imshow(LH_5,[]); title('Ideal LP with D0=5','FontSize',13); i2=i1.*LH_5; i3=ifft2(ifftshift(i2)); subplot(322); imshow(uint8(i3),[]); title('Image of Ideal LP with D0=5','FontSize',13); subplot(323); imshow(LH_50,[]); title('Ideal LP with D0=50','FontSize',13); i2=i1.*LH_50; i3=ifft2(ifftshift(i2)); subplot(324); imshow(uint8(i3),[]); title('Image of Ideal LP with D0==50','FontSize',13); subplot(325); imshow(LH_200,[]); title('Ideal LP with D0=30','FontSize',13); i2=i1.*LH_200; i3=ifft2(ifftshift(i2)); subplot(326); imshow(uint8(i3),[]); title('Image of Ideal LP with D0=200','FontSize',13); figure();

8th Sem - E & C

% %Low pass Butterworth Filter subplot(321); imshow(LHB_5,[]); title('LP Butterworth with D0=5','FontSize',13); i2=i1.*LHB_5; i3=ifft2(ifftshift(i2)); subplot(322); imshow(uint8(i3),[]); title('Image of LP Butterworth with D0=5','FontSize',13); subplot(323); imshow(LHB_50,[]); title('LP Butterworth with D0=50','FontSize',13); i2=i1.*LHB_50; i3=ifft2(ifftshift(i2)); subplot(324); imshow(uint8(i3),[]); title('Image of LP Butterworth with D0=50','FontSize',13); subplot(325); imshow(LHB_200,[]); title('LP Butterworth with D0=200','FontSize',13); i2=i1.*LHB_200; i3=ifft2(ifftshift(i2)); subplot(326); imshow(uint8(i3),[]); title('Image of LP Butterworth with D0=200','FontSize',13); figure();

CGPIT

Department of Electronics & Communication Engineering

Page 35

Fundamental of Image Processing % %Low pass Gausian Filter subplot(321); imshow(LHG_5,[]); title('LP Gausian with D0=5','FontSize',13); i2=i1.*LHG_5; i3=ifft2(ifftshift(i2)); subplot(322); imshow(uint8(i3),[]); title('Image of LP Gausian with D0=5','FontSize',13); subplot(323); imshow(LHG_50,[]); title('LP Gausian with D0=50','FontSize',13); i2=i1.*LHG_50; i3=ifft2(ifftshift(i2)); subplot(324); imshow(uint8(i3),[]); title('Image of LP Gausian with D0=50','FontSize',13); subplot(325); imshow(LHG_200,[]); title('LP Gausian with D0=200','FontSize',13); i2=i1.*LHG_200; i3=ifft2(fftshift(i2)); subplot(326); imshow(uint8(i3),[]); title('Image of LP Gausian with D0=200','FontSize',13); figure(); % %Ideal high pass Filter subplot(321); HH_5=1.-LH_5; imshow(HH_5,[]); title('Ideal HP with D0=5','FontSize',13); i2=i1.*HH_5; i3=ifft2(ifftshift(i2)); subplot(322); imshow(uint8(i3),[]); title('Image of Ideal HP with D0=5','FontSize',13); HH_50=1.-LH_50; subplot(323); imshow(HH_50,[]); title('Ideal HP with D0=50','FontSize',13); i2=i1.*HH_50; i3=ifft2(ifftshift(i2)); subplot(324); imshow(uint8(i3),[]); title('Image of Ideal HP with D0=50','FontSize',13); subplot(325); HH_200=1.-LH_200; imshow(HH_200,[]); title('Ideal HP with D0=200','FontSize',13); i2=i1.*HH_200; i3=ifft2(ifftshift(i2)); subplot(326); imshow(uint8(i3),[]); title('Image of Ideal HP with D0=200','FontSize',13); figure();

8th Sem - E & C

CGPIT

Department of Electronics & Communication Engineering

Page 36

Fundamental of Image Processing

8th Sem - E & C

% %High Pass Butterworth Filter subplot(321); HHB_5=1.-LHB_5; imshow(HHB_5,[]); title('HP Butterworth with D0=5','FontSize',13); i2=i1.*HHB_5; i3=ifft2(ifftshift(i2)); subplot(322); imshow(uint8(i3),[]); title('Image of HP Butterworth with D0=5','FontSize',13); HHB_50=1.-LHB_50; subplot(323); imshow(HHB_50,[]); title('HP Butterworth with D0=50','FontSize',13); i2=i1.*HHB_50; i3=ifft2(ifftshift(i2)); subplot(324); imshow(uint8(i3),[]); title('Image of HP Butterworth with D0=50','FontSize',13); subplot(325); HHB_200=1.-LHB_200; imshow(HHB_200,[]); title('HP Butterworth with D0=200','FontSize',13); i2=i1.*HHB_200; i3=ifft2(ifftshift(i2)); subplot(326); imshow(uint8(i3),[]); title('Image of HP Butterworth with D0=200','FontSize',13); figure(); % %High pass Gaussian Filter subplot(321); HHG_5=1.-LHG_5; imshow(HHG_5,[]); title('HP Gausian with D0=5','FontSize',13); i2=i1.*HHG_5; i3=ifft2(ifftshift(i2)); subplot(322); imshow(uint8(i3),[]); title('Image of HP Gausian with D0=5','FontSize',13); subplot(323); HHG_50=1.-LHG_50; imshow(HHG_50,[]); title('HP Gausian with D0=50','FontSize',13); i2=i1.*HHG_50; i3=ifft2(ifftshift(i2)); subplot(324); imshow(uint8(i3),[]); title('Image of HP Gausian with D0=50','FontSize',13); subplot(325); HHG_200=1.-LHG_200; imshow(HHG_200,[]); title('HP Gausian with D0=200','FontSize',13); i2=i1.*HHG_200; i3=ifft2(fftshift(i2)); subplot(326); imshow(uint8(i3),[]); title('Image of HP Gausian with D0=200','FontSize',13); CGPIT Department of Electronics & Communication Engineering
Page 37

Fundamental of Image Processing

8th Sem - E & C

Result & Conclusion:


Enter Order of Filter : 1 Part: A Ideal Low Pass Filter:

Butterworth Low Pass Filter:

CGPIT

Department of Electronics & Communication Engineering

Page 38

Fundamental of Image Processing Gaussian Low Pass Filter:

8th Sem - E & C

Part: B Ideal High Pass Filter:

CGPIT

Department of Electronics & Communication Engineering

Page 39

Fundamental of Image Processing Butterworth High Pass Filter:

8th Sem - E & C

CGPIT

Department of Electronics & Communication Engineering

Page 40

Fundamental of Image Processing Gaussian High Pass Filter:

8th Sem - E & C

Conclusion:
Hence, we apply the different filtering in frequency domain and observe the result of each on images and we also check results of first and second order filtering.

CGPIT

Department of Electronics & Communication Engineering

Page 41

Fundamental of Image Processing

8th Sem - E & C

Practical No: 8
A. Cyan, Magenta and Yellow B. Hue, Saturation and Intensity C. Luminance(Y),Hue(I) and Saturation(Q) (NTSC format) D. Y, Cb and Cr

Date: 25/03/2013

Aim: Read a full color image and obtain the following color space components;

Program:
clc; clear all; close all; t=imread('C:\Users\Public\Pictures\Sample Pictures\Hydrangeas.jpg'); imshow(t);title('Colour Image'); r=t(:,:,1); g=t(:,:,2); b=t(:,:,3); % %RGB to CMY C = 255-r; M = 255-g; Y = 255-b; figure(); subplot(1,3,1);imshow(C);title('CYAN'); subplot(1,3,2);imshow(M);title('Magenta'); subplot(1,3,3);imshow(Y);title('YELLOW'); figure(); I3=imcomplement(t); subplot(1,2,1);imshow(I3);title('Function CMY'); CMY_Img(:,:,1) = C; CMY_Img(:,:,2) = M; CMY_Img(:,:,3) = Y; subplot(1,2,2);imshow(CMY_Img);title('Equation CMY'); %%RGB to HSV figure(); H=rgb2hsv(t); imshow(H);title('HSV Image');

CGPIT

Department of Electronics & Communication Engineering

Page 42

Fundamental of Image Processing % %RGB to NTSC Y=0.2989 * r + 0.5870 * g + 0.1140 * b; I=0.596*r-0.274*g-0.322*b; Q=0.211*r-0.523*g+0.312*b; figure(); subplot(1,3,1);imshow(Y);title('Luminance'); subplot(1,3,2);imshow(I);title('HUE'); subplot(1,3,3);imshow(Q);title('Intensity'); figure(); I2=rgb2ntsc(t); subplot(1,2,1);imshow(I2);title('Using Function NTSC'); NTSC_Img(:,:,1) = Y; NTSC_Img(:,:,2) = I; NTSC_Img(:,:,3) = Q; subplot(1,2,2); imshow(NTSC_Img);title('Using Equation NTSC'); %RGB to YCbCr R=im2double(r); G=im2double(g); B=im2double(b); Y = 16 + (65.481*R+128.553*G+24.966*B); Cb = 128 + (-37.797*R-74.203*G+112*B); Cr = 128 + (112*R-93.786*G-18.214*B); figure(); subplot(1,3,1);imshow(uint8(Y));title('Y'); subplot(1,3,2);imshow(uint8(Cb));title('Cb'); subplot(1,3,3);imshow(uint8(Cr));title('Cr'); figure(); I4=rgb2ycbcr(t); subplot(1,2,1);imshow(I4);title('Function YCbCr'); YCbCr_Img(:,:,1) = Y; YCbCr_Img(:,:,2) = Cb; YCbCr_Img(:,:,3) = Cr; subplot(1,2,2);imshow(uint8(YCbCr_Img));title('Equation YCbCr');

8th Sem - E & C

CGPIT

Department of Electronics & Communication Engineering

Page 43

Fundamental of Image Processing

8th Sem - E & C

Result & Conclusion:


Original Image:

CGPIT

Department of Electronics & Communication Engineering

Page 44

Fundamental of Image Processing Part: A Cyan, Magenta & Cyan Components of Image:

8th Sem - E & C

CMY Images:

CGPIT

Department of Electronics & Communication Engineering

Page 45

Fundamental of Image Processing Part: B Hue, Saturation & Intensity(HSV) Image:

8th Sem - E & C

Part: C Luminance, Hue & Saturation Components of Image:

CGPIT

Department of Electronics & Communication Engineering

Page 46

Fundamental of Image Processing NTSC Images:

8th Sem - E & C

Part: D Y, Cb & Cr Components of Images:

CGPIT

Department of Electronics & Communication Engineering

Page 47

Fundamental of Image Processing YCbCr Images:

8th Sem - E & C

Conclusion:
Hence, we learn to get the different detail from the colour image and observe the different operation on colour image.

CGPIT

Department of Electronics & Communication Engineering

Page 48

Fundamental of Image Processing

8th Sem - E & C

Practical No: 9 Aim: Read a gray scale image.

Date: 01/04/2013

A. Apply 2 level DWT on image. And Reconstruct image using IDWT B. Apply 1 level DWT and reconstruct image with consider following impact Replace Diagonal Plane element with all zero Replace Horizontal Plane element with all zero Replace Vertical Plane element with all zero Replace Low frequency Plane element with all zero

Find the similarity factor (normalized correlation) for all above case.

Program:
clc;clear all;close all; i=imread('C:\Users\PRADEEP\Desktop\woman.gif'); nbcol = size(i,1); [ll,lh,hl,hh]=dwt2(i,'db10'); subplot(221);imshow(i);title('Original Image'); subplot(222); % % 1-level DWT image representation cod_i = wcodemat(i,nbcol); cod_ll = wcodemat(ll,nbcol); cod_lh = wcodemat(lh,nbcol); cod_hl = wcodemat(hl,nbcol); cod_hh = wcodemat(hh,nbcol); d_1 = [... cod_ll, cod_hl; ... cod_lh, cod_hh ... ]; imshow(d_1,[]); title('1 Level Image','FontSize',13); % % 2-level DWT image representation [llll,lllh,llhl,llhh]=dwt2(ll,'db10'); nbcol = size(ll,2); subplot(223); cod_llll = wcodemat(llll,nbcol); cod_lllh = wcodemat(lllh,nbcol); cod_llhl = wcodemat(llhl,nbcol); CGPIT Department of Electronics & Communication Engineering
Page 49

Fundamental of Image Processing cod_llhh = wcodemat(llhh,nbcol); d1= [... cod_llll, cod_llhl; ... cod_lllh, cod_llhh ... ]; d_resized=imresize(d1,[137 137]); cod_d_resized = wcodemat(d_resized,nbcol); d2 = [... cod_d_resized, cod_hl; ... cod_lh, cod_hh ... ]; imshow(d2,[]); title('2 Level Image','FontSize',13); figure(); % % 2-level IDWT image representation rec_2=idwt2(llll,lllh,llhl,llhh,'db10'); subplot(131); imshow(i);title('Oriignal Image'); subplot(132); imshow(d1,[]); title('2 Level Image','FontSize',13); subplot(133); imshow(rec_2,[]); title('Reconstructed Image from 2level DWT','FontSize',13); figure(); % % 1-level IDWT image representation with hh=0 hh_0=zeros(137,137); rec_1=idwt2(ll,lh,hl,hh_0,'db10'); subplot(131); imshow(i);title('Oriignal Image'); subplot(132); imshow(d1,[]); title('1 Level Image','FontSize',13); subplot(133); imshow(rec_1,[]); title('Reconstructed Image with hh=0','FontSize',13); figure(); % % 1-level IDWT image representation with lh=0 lh_0=zeros(137,137); rec_1=idwt2(ll,lh_0,hl,hh,'db10'); subplot(131); imshow(i);title('Oriignal Image'); subplot(132); imshow(d1,[]); title('1 Level Image','FontSize',13); subplot(133); imshow(rec_1,[]); title('Reconstructed Image with lh=0','FontSize',13); figure(); % % 1-level IDWT image representation with hl=0 hl_0=zeros(137,137); rec_1=idwt2(ll,lh,hl_0,hh,'db10'); CGPIT Department of Electronics & Communication Engineering

8th Sem - E & C

Page 50

Fundamental of Image Processing subplot(131); imshow(i);title('Oriignal Image'); subplot(132); imshow(d1,[]); title('1 Level Image','FontSize',13); subplot(133); imshow(rec_1,[]); title('Reconstructed Image with hl=0','FontSize',13); figure(); % % 1-level IDWT image representation with ll=0 ll_0=zeros(137,137); rec_1=idwt2(ll_0,lh,hl,hh,'db10'); subplot(131); imshow(i);title('Oriignal Image'); subplot(132); imshow(d1,[]); title('1 Level Image','FontSize',13); subplot(133); imshow(rec_1,[]); title('Reconstructed Image with ll=0','FontSize',13);

8th Sem - E & C

Results & Conclusion:


Part: A 1 level & 2 level DWT of given Image:

CGPIT

Department of Electronics & Communication Engineering

Page 51

Fundamental of Image Processing 2 level DWT image & reconstucted image from it:

8th Sem - E & C

Part: B

1 level DWT & reconstucted image with HH=0:

CGPIT

Department of Electronics & Communication Engineering

Page 52

Fundamental of Image Processing 1 level DWT & reconstucted image with LH=0:

8th Sem - E & C

1 level DWT & reconstucted image with HL=0:

CGPIT

Department of Electronics & Communication Engineering

Page 53

Fundamental of Image Processing 1 level DWT & reconstucted image with LL=0:

8th Sem - E & C

Conclusion:
After completion of this practical we observe that in DWT the LL part contain almost all information of the original image, LH part contain horizontal and HL contain vertical information, HH part contain diagonal information. Now, if we remove the HH, LH or HL part from the DWT image almost all information is present in the image but if we remove LL part, very less detail of image remains.

CGPIT

Department of Electronics & Communication Engineering

Page 54

Fundamental of Image Processing

8th Sem - E & C

Practical No: 10 Aim: Read a gray scale image. Convert it into 4 gray level (2-bit) image.
Find out the probability of gray level (0,1,2,3) Find out Arithmetic code for the input sequence: 20131.

Date: 08/04/2014

Program:
clc;clear all;close all; i=imread('cameraman.tif'); subplot(131) imshow(i);title('Original Image'); subplot(132); i=double(i); i1=floor(i./64); [m,n]=size(i1); p=zeros(4,1); for j=1:m for k=1:n; for l=1:4 if i1(j,k)==l-1; p(l,1)=p(l,1)+1; end end end end imshow(i1,[]);title('2 Bit Image'); subplot(133); s=p/(m*n); bar(s);title('Histogram'); s1=sort(s); % % CDF for 1st step r11=s1(1,:); r12=r11+s1(2,:); r13=r12+s1(3,:); r14=r13+s1(4,:); r1=[r11 r12 r13 r14]' % % CDF for 2nd step for '2' r_21=r12; dif=r13-r12; r_22=r12+(dif*r12); r_23=r12+(dif*r13); r_24=r12+(dif*r14); CGPIT Department of Electronics & Communication Engineering
Page 55

Fundamental of Image Processing r_2=[r_21 r_22 r_23 r_24]' % % CDF for 3rd step for '0' r_01=r_21; dif=r_22-r_21; r_02=r_21+(dif*r12); r_03=r_21+(dif*r13); r_04=r_21+(dif*r14); r_0=[r_01 r_02 r_03 r_04]' % % CDF for 4th step for '1' r_11=0; dif=r_01-0; r_12=(dif*r12); r_13=(dif*r13); r_14=(dif*r14); r_1=[r_11 r_12 r_13 r_14]'

8th Sem - E & C

% % CDF for 5th step for '3' r_31=r_13; dif=r_14-r_13; r_32=r_13+(dif*r12); r_33=r_13+(dif*r13); r_34=r_13+(dif*r14); r_3=[r_31 r_32 r_33 r_34]' % % CDF for final step for '1' r_1_1=r_31; dif=r_32-r_31; r_1_2=r_31+(dif*r12); r_1_3=r_31+(dif*r13); r_1_4=r_31+(dif*r14); r_11=[r_1_1 r_1_2 r_1_3 r_1_4]'

CGPIT

Department of Electronics & Communication Engineering

Page 56

Fundamental of Image Processing

8th Sem - E & C

Results & Conclusion:


Original & 2-bit Image with Histogram:

Output on Command Window: r1 = 0.0179 0.1773 0.4219 1.0000 r_2 = 0.1773 0.2206 0.2805 0.4219 CGPIT Department of Electronics & Communication Engineering
Page 57

Fundamental of Image Processing r_0 = 0.1773 0.1850 0.1956 0.2206 r_1 = 0 0.0314 0.0748 0.1773 r_3 = 0.0748 0.0930 0.1180 0.1773 r_11 = 0.0748 0.0780 0.0825 0.0930

8th Sem - E & C

Conclusion:
Hence, we convert the gray scale image in to 2-bit image using the thresholding and observe the result and we conclude that this image contain almost all the information of real image but contain less memory space. We also calculated the probability of that image and generate the arithmetic code for given sequence.

CGPIT

Department of Electronics & Communication Engineering

Page 58

Das könnte Ihnen auch gefallen