Digital Image Processing using Scilab

R.Senthilkumar,Assistant Professor,

Department of Electronics and Communication Engineering,

Institute of Road and Transport Technology,Erode – 638 316

[1]. Image Arithmetic//Prog1.Image Arithmetic - To learn to use arithmetic operations to combine images. [Program]

clc;

clear;

close;

I = imread('C:\Users\Senthil\Desktop\DigitalImageProcessing\cameraman.jpeg'); //SIVP toolbox

J = imread('C:\Users\Senthil\Desktop\DigitalImageProcessing\rice.png');//SIVP toolbox

//Addition of two images

IMA = imadd(I,J); //SIVP toolbox

figure

ShowImage(IMA,'Image Addition')//IPD toolbox

//Difference between two images

IMS = imabsdiff(I,J);//SIVP toolbox

figure

ShowImage(IMS,'Image Subtraction');//IPD toolbox

//Division of two images

IMD = imdivide(I,J);//SIVP toolbox

IMD = imdivide(IMD,0.01);//SIVP toolbox

figure

ShowImage(uint8(IMD),'Image Division');//IPD toolbox

//Multiplication of two images

IMM = immultiply(I,I);//SIVP toolbox

figure

ShowImage(uint8(IMM),'Image Multiply');//IPD toolbox

Image Addition

Image Division

Image Subtraction

Image Multiplication

[2]. Image Format Conversion (RGB to Gray and Gray to Binary)

//Program: Image Format Conversion [Program]//[1].RGB to Gray

//[2]. Gray to Binary

clc;

clear;

close;

I = imread('C:\Users\Senthil\Desktop\DigitalImageProcessing\redrose.jpg');

ShowColorImage(I,'Red Rose Color Image')

//RGB to Gray Image

I_gray = rgb2gray(I);

figure

ShowImage(I_gray,'Red Rose Gray Image')

Imean = mean2(I_gray);

Ithreshold = double(Imean)/double(max(I_gray(:)));

I_bw = im2bw(I,Ithreshold);

figure

ShowImage(I_bw,'Red Rose Binary Image')

Red Rose Colour Image Red Rose Gray Image

Red Rose Binary Image

[3]. Image Interpolation//To learn the role of interpolation operation

// i) Bi-linear ii) Bi-cubic iii) nearest neighbor [Program]

clc;

clear;

close;

I = imread('C:\Users\Senthil\Desktop\DigitalImageProcessing\lenna512.png');//size 512x512

ShowColorImage(I,'Lenna512 Colour Image')

title('Lenna512 Colour Image')

I = rgb2gray(I);

[m,n] = size(I);

I_nearest = imresize(I,[256,256]); //'nearest' - nearest-neigbor interpolation -SIVP Atom

I_bilinear = imresize(I,[256,256],'bilinear');// 'bilinear' - bilinear interpolation

I_bicubic = imresize(I,[256,256],'bicubic');//'bicubic' - bicubic interpolation

figure

ShowImage(uint8(I_nearest),'nearest-neigbor interpolation');

title('nearest-neigbor interpolation')

figure

ShowImage(uint8(I_bilinear),'bilinear - bilinear interpolation');

title('bilinear - bilinear interpolation')

figure

ShowImage(uint8(I_bicubic),'bicubic - bicubic interpolation');

title('bicubic - bicubic interpolation')

[4]. Point Operations- Image Enhancement

//Point Operations - To learn image enhancement through point transformation [Program]

//i)Linear transformation ii) Non-linear transformation

clc;

clear;

close;

I = imread('C:\Users\Senthil\Desktop\DigitalImageProcessing\rice.png');

//(i). Linear Transformation

//IMAGE NEGATIVE

I = double(I);

J = 255-I;

figure

ShowImage(I,'Original Image')

title('Original Image')

figure

ShowImage(J,'Linear Transformation-IMAGE NEGATIVE')

title('Linear Transformation-IMAGE NEGATIVE')

//(ii) Non-linear transformation

//GAMMA TRANSFORMATION

GAMMA = 0.9;

K = I.^GAMMA;

figure

ShowImage(K,'Non-linear transformation-GAMMA TRANSFORMATION')

title('Non-linear transformation-GAMMA TRANSFORMATION')

[5]. Image Histogram Plot

//Caption: Histogram Plot [Program]

clc;

clear;

close;

a=imread('C:\Users\Senthil\Desktop\DigitalImageProcessing\pout.jpg')

[m n]=size(a);

for i=1:256

b(i)=length(find(a==(i-1)));

end

figure,

ShowImage(a,'Original Image')//IPD toolbox

title('Original Image')

figure

plot2d3('gnn',[0:255],b)

title('Histogram of the Image')

//using SIVP Atom

[count,cells]=imhist(a);

figure

plot2d3('gnn',cells,count)

title('Histogram plot using SIVP atom')

//using IPD Atom

Histogram = CreateHistogram(a,cells);

figure

plot2d3('gnn',cells,Histogram)

title('Histogram plot using IPD atom')

[6]. Histogram Equalization

//Image Histogram Equalisation[Program]

clc;

close;

f=ReadImage('C:\Users\Senthil\Desktop\DigitalImageProcessing\pout.jpg');// Get image (Keep image file in same folder)

OrigSize=size(f);//Get size of original image

OrigRow=OrigSize(1);//Get number of rows in original image

OrigCol=OrigSize(2);//Get number of colomns in original image

figure

ShowImage(f,'Original Image Pout')

WriteImage(f,'Exp6OriginalImagePout.jpg') //Comment out if you do not want o/p file to be written on disk

OrigHist=CreateHistogram(uint8(f));//Create Histogram

figure

plot2d3(OrigHist)//Show Histogram

title('Original Image Histogram')

P=OrigHist/(OrigRow*OrigCol);//get average pixel value

C=zeros(1,256);

for i=2:256

C(1,i)=C(1,i-1)+P(1,i);

end

●

Cdash = round(C*255);

EqImage=[]//Empty matrix for output image

for i=1:OrigRow

for j=1:OrigCol

EqImage(i,j)=Cdash(1,f(i,j)+1);

end

end

figure

ShowImage(uint8(EqImage),'Image Histogram Equalisation')

WriteImage(uint8(EqImage),'Exp6ImageHistogramEqualisation.jpg') //Comment out if you do not want o/p file to be written on disk

EqHist=CreateHistogram(uint8(EqImage));//Create Histogram

figure

plot2d3(EqHist)//Show Histogram

title('Histogram Equalized Image')

[7]. Morphological Operations: Dilation and Erosion [Program]

clc;

clear;

close;

Image = imread('C:\Users\Senthil\Desktop\DigitalImageProcessing\tire.jpeg');

StructureElement = CreateStructureElement('square',3); // generate structuring element IPD atom

ResultImage1 = ErodeImage(Image,StructureElement); //IPD Atom

ResultImage2 = DilateImage(Image,StructureElement); //IPD Atom

ResultImage3 = BottomHat(Image,StructureElement); //IPD Atom

ResultImage4 = TopHat(Image,StructureElement); //IPD Atom

figure

ShowImage(Image,'Original Image')

title('Original tire.jpg Image')

figure

ShowImage(ResultImage1,'Eroded Image')

title('Eroded Image')

figure

ShowImage(ResultImage2,'Dilated Image')

title("Dilated Image")

figure

ShowImage(ResultImage3,'bottom hat filtered image')

title('Botton hat filtered Image')

figure

ShowImage(ResultImage4,'top hat filtered image')

title('Top hat Filtered Image')

ResultImage5 = imadd(ResultImage3,ResultImage4);

figure

ShowImage(ResultImage4,'top hat filtered image+bottom hat filtered image')

title('Top hat filtered image+ Bottom hat Filtered Image')

[8]. Colour Image Manipulations [Program]

clc

clear

close

//Showing RGB components of a color RGB image.

//Splitting the color image (RGB Image) into three planes

a=imread('C:\Users\Senthil\Desktop\DigitalImageProcessing\peppers.png'); //this image is 348x512x3 size

figure

ar=a(:,:,1);

ShowImage(ar,'RED Matrix')

figure

ag=a(:,:,2);

ShowImage(ag,'GREEN Matrix')

figure

ab=a(:,:,3);

ShowImage(ab,'BLUE Matrix')

RGB = imread('C:\Users\Senthil\Desktop\DigitalImageProcessing\peppers.png');//SIVP toolbox

RGB_128 = RGB/2;

RGB_128 = round(RGB_128)

[X,map] = RGB2Ind(RGB_128);

figure

ShowImage(X,'Indexed Image',map)

figure

ShowColorImage(RGB,'RGB Color Image')

YIQ = rgb2ntsc(RGB);

figure

ShowColorImage(YIQ,'NTSC image YIQ')

RGB = ntsc2rgb(YIQ);

YCC = rgb2ycbcr(RGB);

figure

ShowColorImage(YCC,'equivalent HSV image YCbCr')

RGB = ycbcr2rgb(YCC);

HSV = rgb2hsv(RGB);

figure

ShowColorImage(HSV,'equivalent HSV image')

RGB = hsv2rgb(HSV);

[9]. Fourier Transform of 2D Gray Image [Program]

//Caption: 2D-Discrete Fourier Transform using inbuilt fast fourier transform

//function fft2()

clc;

clear;

close;

a = ReadImage('C:\Users\Senthil\Desktop\DigitalImageProcessing\lenna.jpg');

[m,n]= size(a);

ShowImage(a,'Original lenna Image');

title('Original lenna Image')

//2D-DFT using FFT

A = fft2(double(a));

figure(1)

ShowImage(abs(A),'2D-Discrete Fourier Transform of Lena Image');

title('2D-Discrete Fourier Transform of Lena Image')

//fftshifted image

B = fftshift(A)

figure(2)

ShowImage(abs(B),'2D-Discrete Fourier Transform of Lena Image-fftshifted');

title('2D-Discrete Fourier Transform of Lena Image-fftshifted')

//2D-IDFT using FFT

a_inv = fft2(A')

a_inv = a_inv'/(m*n);

figure(3)

ShowImage(uint8(abs(a_inv)),'2D-Inverse Discrete Fourier Transform ');

title('2D-Inverse Discrete Fourier Transform')

[10]. Discrete Wavelet Transform and Inverse Discrete Wavelet Transform [Program]

//Caption: Scilab code to implement Discrete Wavelet Transform

clc;

clear;

close;

//Original Image

img = imread('C:\Users\Senthil\Desktop\DigitalImageProcessing\cameraman.jpeg');

figure(1)

ShowImage(img,'Original Image')

title('Original Image');

[p q] = size(img);

[CA,CH,CV,CD] = dwt2(double(img),'db1');//Compute 2D wavelet transform -daubcheis wavelet

figure(2)

ShowImage(uint8(CA),'LPF-LPF output')

title('LPF-LPF output')

figure(3)

ShowImage(uint8(CH),'LPF-HPF output')

title('LPF-HPF output')

figure(4)

ShowImage(uint8(CV),'HPF-LPF output')

title('HPF-LPF output')

figure(5)

ShowImage(uint8(CD),'HPF-HPF output')

title('HPF-HPF output')

img_inv = idwt2(CA,CH,CV,CD,'db1');

img_inv = uint8(img_inv);

figure(6)

ShowImage(img_inv,'Reconstructed Image')

title('Reconstructed Image')

[11]. Edge Detection [Program]

//Caption: Scilab code for Edge Detection using Different Edge detectors

//[1]. Sobel [2].Prewitt

close;

a = imread('C:\Users\Senthil\Desktop\DigitalImageProcessing\lenna.jpg');

c = edge(a,'sobel');

d = edge(a,'prewitt');

ShowImage(a,'Original Image')

title('Original Image')

figure

ShowImage(c,'Edge Detected Image-Sobel')

title('Edge Detected Image -Sobel')

figure

ShowImage(d,'Edge Detected Image-Prewitt')

title('Edge Detected Image-Prewitt')

//Caption: Scilab code for Edge Detection using Different Edge detectors [Program]

//[1]LoG [2] DoG

close;

a = imread('C:\Users\Senthil\Desktop\DigitalImageProcessing\lenna.jpg');

//LAPLACIAN OF GAUSSIAN////////////////////////////////////////////////

[c,thresh] = edge(a,'log',0.2,'both',2);

//d = edge(a,'prewitt');

ShowImage(a,'Original Image')

title('Original Image')

figure

ShowImage(c,'Edge Detected Image-LoG')

//DERIVATIVE OF GAUSSIAN FUNCTION

function [mask]=DoG(op1,op2)

//If op1 and op2 are not specified -Default size 3x3

if isempty(op1) then

siz = [3,3];

else

if length(op1)==1 then

siz = [op1, op1];

elseif length(op1)==2 then

siz = op1;

else

error("The second argument should have 1 or 2 elements for gaussian filter");

end

end

//set std for the filter

if isempty(op2) then

g_std = 0.5;

else

if length(op2)>1 then

error("The third argument should have only 1 element for gaussian filter");

else

g_std = op2;

end

end

sizx = (siz(2)-1)/2;

sizy = (siz(1)-1)/2;

x2 = ones(siz(1),1) * ([-sizx:sizx]^2);

y2 = ([-sizy:sizy]^2)' * ones(1, siz(2));

r = sqrt(x2+y2);

sigma = g_std;

F = (1/(sigma^2))*(((r.*r)/sigma^2)-1).*exp(-r.*r/2*sigma^2);

F(F<%eps*max(F)) = 0;

sumF=sum(F);

if sumF~=0 then

F = F / sum(F);

end

mask = F;

endfunction

mask = DoG([ ],[ ]);

mx=filter2(mask,a);

my=filter2(mask',a);

border=sqrt(mx.*mx + my.*my);

if thresh >=0 then

scale_thresh = min(border) * (1-thresh) + max(border)*thresh;

border=border > scale_thresh;

end

d = border;

figure

ShowImage(d,'Edge Detected Image-DoG')

Image Enhancement Methods

[13].Brightness enhancement

[14]. Brightness suppression

[15].Contrast Manipulation

[16]. Image Negative

[17]. Threshold Operation on Gray Image

[18]. Gray level slicing without background

[19]. Image Cropping, Image Complement and Linear Combination of Images

//Image Arithmetic: Image Complement, Image //Cropping and Linear Combination of Images [Program]

clc;

clear;

close;

I = imread('C:\Users\senthilkumar\Desktop\Gautam_PAL_Lab\DIP_Lab2\cameraman.jpeg');

J = imread('C:\Users\senthilkumar\Desktop\Gautam_PAL_Lab\DIP_Lab2\lenna.jpg');

K = imabsdiff(I,J);

ShowImage(I,'Cameraman Image')

title('Cameraman Image')

figure

ShowImage(J,'Lenna Image')

title("Lenna Image")

figure

ShowImage(K,'Absolute Difference Between cameraman and Lenna Image')

title('Absolute Difference Between cameraman and Lenna Image')

L = imcomplement(K);

figure

ShowImage(L,'Complement of difference Image K ')

title('Complement of difference Image ')

rect = [20,30,200,200];

I_subimage = imcrop(I,rect);

J_subimage = imcrop(J,rect);

figure

ShowImage(I_subimage,'Sub Image of Cameraman Image')

title('Sub Image of Cameraman Image')

figure

ShowImage(J_subimage,'Sub Image of Lenna Image')

title('Sub Image of Lenna Image')

a=2;

b =0.5;

M = imlincomb(a,I,b,J);

figure

ShowImage(M,'Linear Combination of cameraman and Lenna Image')

title('Linear Combination of cameraman and Lenna Image')

N= imlincomb(b,I,a,J);

figure

ShowImage(N,'Linear Combination of cameraman and Lenna Image')

title('Linear Combination of cameraman and Lenna Image')

[20]. Image Filtering and De-noising

//Neighborhood Operations - To learn about neighborhood operations and use them for

//i) Linear filtering ii) Non-linear filtering [Program]

clc;

clear;

close;

I = imread('C:\Users\Senthil\Desktop\DigitalImageProcessing\lenna.jpg');

I_noise = imnoise(I,'salt & pepper');

figure

ShowImage(I,'Original Lenna Image')

title('Original Lenna Image')

figure

ShowImage(I_noise,'Noisy Lenna Image')

title('Noisy Lenna Image')

//Case 1: Linear Filtering

//(i).Linear Filtering -Example 1: Average Filter

F_linear1 = 1/25*ones(5,5);//5x5 mask

I_linear1 = imfilter(I_noise,F_linear1);//linear filtering -Average Filter

figure

ShowImage(I_linear1,'Linear Average Filtered Noisy Lenna Image')

title('Linear Average Filtered Noisy Lenna Image')

//(ii).Linear Filtering - Example 2: Gaussing filter

hsize = [5,5];

sigma = 1;

F_linear2 = fspecial('gaussian', hsize, sigma); //Linear filtering- gaussian Filter

I_linear2 = imfilter(I_noise,F_linear2);

figure

ShowImage(I_linear2,'Linear Gaussian Filtered Noisy Lenna Image')

title('Linear Gaussian Filtered Noisy Lenna Image')

//Case 2: Non-Linear Filtering (i).Median Filtering

F_NonLinear = [3,3];

I_NonLinear = MedianFilter(I_noise,F_NonLinear);//Median Filter 3x3

figure

ShowImage(I_NonLinear,'Median Filtered(Non-Linear) Noisy Lenna Image')

title('Median Filtered(Non-Linear) Noisy Lenna Image')

Video Processing -Introduction

Scilab Image and Video Processing [Program]

//Program Simple video reading and writing .avi formats and manipulation of video frames.

//Note 1: Install xvid codec for read and write video files from

//http://www.xvid.org/Downloads.15.0.html

//Note 2: very large video can not be read by scilab

//Note 3: shuttle.avi is a large file more 100 frames. use shuttlenew.avi file

//for video processing applications

//Using SIVP Atom

//Software version

//OS Windows7 ,8

//Scilab5.4.1 and above

//Image Processing Design Toolbox 8.3.1-1

//Scilab Image and Video Proccessing toolbox 0.5.3.1-2

clear;

clc;

close;

n = aviopen('C:\Users\Senthil\Desktop\DigitalImageProcessing\shuttle_Xvid.avi');

im = avireadframe(n); //get a frame

imshow(im);

avilistopened()

aviclose(n);

Image Processing Design Atom is a better tool for Scilab video Processing [Program]

clear;

clc;

close;

VideoPath = 'C:\Users\Senthil\Desktop\DigitalImageProcessing\shuttle_Xvid.avi';

VideoInfo = GetVideoStruct('C:\Users\Senthil\Desktop\DigitalImageProcessing\shuttle_Xvid.avi');

VideoFilePointer = OpenVideoFile('C:\Users\Senthil\Desktop\DigitalImageProcessing\shuttle_Xvid.avi');

figure();

for n = 1 : VideoInfo.NumberOfFrames

RGB = ReadImage(VideoFilePointer);

ShowColorImage(RGB, VideoPath);

end;

CloseVideoFile(VideoFilePointer);

VideoInfo =

Codec: "xvid"

NumberOfFrames: 121

FrameRate: 30.00003

Width: 512

Height: 288

IsColor: %t