HOW TO CONVERT GRAY IMAGE TO HUE IMAGE?

i am trying to find psnr value of hue image and gray image

i need to know how to convert gray image to hue

Answer 1

[KALYAN ACHARJYA]

Have you read about Hue? Read here in wiki

I dont think there is any conversiotion way, if you do somehow, all gray pixels=0 in hue space (color appearance parameters).

Must read thread here

[tashu Dabariya]

sir i have an another question how to convert binary to hue?

[John D'Errico]

A binary image to hue? That would seem to be meaningless. A binary image has only TWO colors, black and white. Unless you supply additional information, what hues would you want to see?

[tashu Dabariya]

then sir how to find psnr of hue image and the binary image?

[tashu Dabariya]

img=imread('.jpg');

% imresize for resize image

img=imresize(img,[256 256]);

img=im2double(img);

figure,imshow(img),title('original image');

% convert rgb to hsv

img1=rgb2hsv(img);

H=img1(:,:,1);

S=img1(:,:,2);

V=img1(:,:,3);

H(:,:,1) = H(:,:,1) * 2.5; %increase hue image

figure(), subplot(2, 2, 1), imshow(img), title('Original')

subplot(2, 2, 2), imshow(H), title('Hue'), colorbar

subplot(2, 2, 3), imshow(S), title('Saturation'), colorbar

subplot(2, 2, 4), imshow(V), title('Value of Brightness'), colorbar

% otsu thresholding

fim=rgb2gray(img);

level=graythresh(fim);

bwfim=im2bw(fim,level);

figure(),imshow(bwfim),title('Otsu')

sir, I want to find accuracy through psnr and mse or any option to find the accuracy of hue image and binary image (Otsu)

I don't understand how to find the accuracy of hue and binary image

[tashu Dabariya]

H = (H);

bwfim =double (bwfim);

% Find the mean squared error

mse = sum((H(:)-bwfim(:)).^2) / numel(H)

% now find the psnr, as peak=255

psnr = 10*log10(255*255/mse)

want to know is it right code to find mse and psnr?

Answer 2

[Image Analyst]

The code you gave is irrelevant. It will produce bogus results. To compare colors you need to use Delta E. Plus, it's meaningless to compare the hue channel of a color image to a binary image. You need to describe what you want better so I know how to answer you. I do color image analysis every day (for decades), so I might know a little about it.

[tashu Dabariya]

I want to identify greenness in plant

1st convert original image (RGB) to HSV used hue image because of the illumination conditions (my question is which value find psnr, original image or hue image?)

and last compare with ExG, Otsu(these images converted into binary) and the last question is the same (which value is found psnr binary image or ExG Otsu original image?)

[tashu Dabariya]

clear variables

close all

clc

%

img=imread('1.jpg');

img=imresize(img,[256 256]);

img=im2double(img);

figure(),imshow(img),title('Original Image');

%Convert Rgb to hsv

img1=rgb2hsv(img);

H=img1(:,:,1);

S=img1(:,:,2);

V=img1(:,:,3);

%

H(:,:,1) = H(:,:,1) * 2.5;

%

% rgbImage = hsv2rgb(img1);

figure(), subplot(2, 2, 1), imshow(img), title('Original')

subplot(2, 2, 2), imshow(H), title('Hue'), colorbar

subplot(2, 2, 3), imshow(S), title('Saturation'), colorbar

subplot(2, 2, 4), imshow(V), title('Value of Brightness'), colorbar

% visible spectral index methods

R =img(:,:,1);

G = img(:,:,2);

B = img(:,:,3);

g1r = R./(R+G+B);

g1g = G./(R+G+B);

g1b = B./(R+G+B);

ExG = 2.*g1g-g1r-g1b;

figure(),subplot(2,7,1); imagesc(ExG);title(' ExG greenness')

% Determine max, min, and delta.

[imageMax,imageMaxIndex] =(max(img,[],3));

imageMin = min(img,[],3);

imageDelta = imageMax - imageMin;

% Calculate the hue.

imageMaxRed = imageMaxIndex==1;

imageMaxGreen = imageMaxIndex==2;

imageMaxBlue = imageMaxIndex==3;

% Calculate the hue

H = zeros(size(img1(:,:,1)));

H(imageMaxRed) =((G(imageMaxRed)-B(imageMaxRed)))./(imageDelta(imageMaxRed))./6;

H(imageMaxGreen) = (2+(B(imageMaxGreen)- R(imageMaxGreen)))./(imageDelta(imageMaxGreen))./6;

H(imageMaxBlue) = (4+(R(imageMaxBlue)-G(imageMaxBlue)))./(imageDelta(imageMaxBlue))./6;

% Perform thresholding on the hue image.

imageHueThresh = graythresh(H);

imageHueBw = im2bw(H,imageHueThresh);

%Remove Background

R=0;

if (H<50)|(H>150)

else

R=R;

end

G=0;

if H<50|H>150

else H=H;

end

B=0;

if (H<50)|(H>150)

else

B=B;

end

% Greenness Identification using ExG

T = 49; %T can be any value in [1, 49]

imageHueThresh = T./255;

ExG = 2.*g1g-g1r-g1b;

imageHueBw=1;

if ExG > T;

else

imageHueBw=0;

end

imageHueBw=im2bw(ExG,imageHueThresh);

figure(), imshow(imageHueBw),title('Greenness Identification using ExG')

% %delete the obejcts less than 100 pixels

% imageHueBw=bwareaopen(imageHueBw,100);

% figure,imshow(imageHueBw);

%Otsu Thresholding

fim=rgb2gray(img);

level=graythresh(fim);

bwfim=im2bw(fim,level);

figure(),imshow(bwfim),title('Otsu');

this is my code i want to know this code is right to identify greenness nd want to know how to identify psnr ?

[Image Analyst]

Use the Color Thresholder app on the Apps tab of the tool ribbon. Choose HSV color space. Adjust the thresholds until you get green. Then click the Export code button to get a function that will do it.

For the PSNR, what is your reference image?

[tashu Dabariya]

imageHueBw=im2bw(ExG,imageHueThresh);

bwfim=im2bw(fim,level);

imageHueBw and bwfim is my refrence image both of these are binary image

i am just asked for psnr which image is used original image or hue image?

and my refrence image is binary how to get psnr value?

plss help

[tashu Dabariya]

i want accuracy of identification of greenness

how to find accuracy of identification?

[Image Analyst]

None of what you said about PSNR makes sense.

To find accuracy of identification, you might use your binary image of what your algorithm found for "green" and compare it to what you hand drew with imfreehand(). Then use the Sorensen Dice coefficient (see Wikipedia).