Image Intensity

Hi all, Can any one tel me the command in matlab to calculate the intensity of image.

[Jean]

Does anyone further know how to do this for a specific area of pixels? For example, image(100:1:200,100:1:200) for a 100 by 100 pixel area, and get the intensity in this area as an average?

[Image Analyst]

meanGrayLevel = mean2(yourImage(100:200,100:200));

Don't use "image" as the name of your variable since it's a function.

Answer 1

[Image Analyst]

The mean of a particular pixel is just simply the pixel value, since there is only one sample - nothing really to take the mean of:

intensityValue = grayImage(100, 230); 
meanIntensityValue = mean(grayImage(100, 230));

Here, of course meanIntensityValue will equal intensityValue.

To get the mean of all the pixel values in the entire image, you can do any of these:

meanIntensityValue = mean2(grayImage);
meanIntensityValue = mean(grayImage(:));
meanIntensityValue = mean(mean(grayImage));

[M@lik Ali]

thanks for the response

Can we calculate the intensity like this

grayImage = rgb2gray(I);

[M N d] = size(grayImage);

intensityValue=0;

count=0;

for x = 1:M

for y = 1:N

intensityValue = intensityValue + uint32(grayImage(x,y)) ;

end;

end;

[Image Analyst]

Yes, if you divide that by M*N, and if you want to do it much much slower than my method.

[Mona]

I want the average of the intensities of all pixels. does this work on 3D hyperscectral image?

meanIntensityValue = = mean(HyperImage(:));

[Mona]

or

meanIntensityValue = mean2(HyperImage);

[Iain]

Yes. Both should work. (Though, you have a spare "=" in the first one).

[Mona]

Thank you so much

[Image Analyst]

Please mark the discussion as "Answered" if you're done with it. Unless you want to open a can of confusing worms about how the units of an image are not really intensity.

[Samantha Villanueva]

hello can I use this to a colored image?

[Image Analyst]

You would have to do it on every color channel independently:

[r, g, b] = imsplit(rgbImage);
meanRed = mean2(r);
meanGreen = mean2(g);
meanBlue = mean2(b);

[Walter Roberson]

As the question is about mean intensity, then which definition of "intensity" do you want to use for your RGB image?

[Jitin Malhotra]

@Image Analyst,@Walter Roberson,

verticalProfile = sum(testimage, 2);
figure; barh(verticalProfile);

so i put a threshold on the sum from vertical profile say at sum of 70000, removing all pixels based on this sum. Then what would be the units of this threshold value 70000? Will i call this pixels or brightness value, intensity value, what would i call them?

[Walter Roberson]

If you are using a grayscale image in which the values are "brightness" (as opposed to something like Hue), then the real unit would be closer to "brightness per pixel^2". Summing that over a number of pixels in a row or column would give you "brightness per pixel^2 * pixels" which would be "brightness per pixel".

... Brightness would not be the real physical unit. Eyes do not perceive brightness in a linear manner. https://www.pathwaylighting.com/products/downloads/brochure/technical_materials_1466797044_Linear+vs+Logarithmic+Dimming+White+Paper.pdf

"Approximate perceived brightness = sqrt(measured brightness)"

I just woke up, and digging into luminance and chromaticities and transfer functions this early in my day is making my brain hurt. @Image Analyst studied those things, so he would be in a much better position to talk about what the real physical units would be.

[Image Analyst]

The usual nomenclature of your vertical profile (the image summed horizontally across columns) is "integrated gray value" and it will have units of gray levels. For a standard digital camera which measures illuminance Lux or irradiance (watts per square meter), the units of gray level correspond to units of energy (Joules).

(watts/m^2) * (area of pixel in m^2) * (seconds of exposure time) = joules.

Note that watts = joules/second so everything cancels out except for joules.

Note: units prefixed with "Phot" or "Lum" (like illuminance) refer to radiation in the visible range 400-700 nm. If it doesn't have a prefix or has "rad" prefix (like irradiance), then it applies to all energy across the whole electromagnetic spectrum.

Answer 2

[swetha Derangula]

Is this code correct for entire image intensity? I didn't get any intensity value

clear all;
close all;
o=imread('450.jpg');
i=rgb2gray(o);
imshow(i)
intensityValue = i(100, 230);  
meanIntensityValue = mean(mean(i));

please suggets me

[Walter Roberson]

No code that contains "clear all" can be correct.

Your line

intensityValue = i(100, 230);

extracts one particular pixel's intensity, and only that one pixel. However,

meanIntensityValue = mean(mean(i));

should be fine for taking the mean intensity of the entire image.

[Image Analyst]

You could also use mean2().

Answer 3

[Adina Stoica]

Well, the intensity of the image is the average of the intensities of all pixels in an image (unless you're doing anything more specific than that). So if you do

mean(imagefile(:))

you should get the intensity of the image.

[M@lik Ali]

Adina thanks for your response,

But actually i want to calculate the intensity value of all elements of the images.

someone told me like

intensityValue = grayImage(100, 230);

But in i have to calculate the intensity value of each pixel or total image, i think i can do it by calculating the intensity of each pixel and then add it but i am not sure.

Kindly help me about it..

Answer 4

[Saranya]

hi image analyst i want to measure the intensity profile for gray scale image.. can you help me..?

[Image Analyst]

You can use the improfile() function to do that.

[Saranya]

clc; clear all; close all;

I used this command sir but its not working.. this is my input image http://postimg.org/image/pe5ozahzb/e770bd0c/

[f p] = uigetfile('*.jpg');
I = imread([p f]);
J = rgb2gray(I);
improfile()

[Image Analyst]

You're not taking the result of improfile. What variable to you want to use to accept the intensity numbers from improfile()?

lineProfile = improfile().

[Saranya]

still am not getting the answer sir..

[f p] = uigetfile('*.jpg');
I = imread([p f]);
J = rgb2gray(I);
c = improfile();

I need a plot with X and Y axis. so that the lowest intensity will be the centromere position

[Image Analyst]

Well you never plotted it! You need to call plot(). Try this:

fontSize = 20;
folder = fullfile(matlabroot, '\toolbox\images\imdemos');
filePattern = fullfile(folder, '*.jpg')
[f p] = uigetfile(filePattern);
subplot(2, 1, 1);
originalImage = imread([p f]);
grayImage = rgb2gray(originalImage);
imshow(grayImage);
title('Original Image', 'FontSize', fontSize);
axis on;
% Enlarge figure to full screen.
set(gcf, 'Units', 'Normalized', 'OuterPosition', [0 0 1 1]);
% Give a name to the title bar.
set(gcf, 'Name', 'Demo by ImageAnalyst', 'NumberTitle', 'Off')

uiwait(msgbox('Left click, then right click'));
c = improfile();
subplot(2, 1, 2);
plot(c, 'LineWidth', 3);
grid on;
title('Intensity Profile', 'FontSize', fontSize);

[Saranya]

Thank you very much sir. I got the result.

[Saranya]

Sir, Is there any way to get the intensity profile automatically. i.e., without drawing the line manually?

[Image Analyst]

Yes. improfile() can take arguments that are the endpoints of the line.

[Jensa]

Hi Image Analyst...Can you help me to find the density profile for my chromosome image..Attaching the image along with this.. http://postimg.org/image/pe5ozahzb/e770bd0c/

[Image Analyst]

Please start your own discusssion.

[Rasika Devi]

hello image analyst. i want to know how to calculate the weight of an image?

[Walter Roberson]

Rasika Devi, could you confirm that you want to be able to calculate the number of photons that would be emitted to display a particular image for a particular length of time, and you would want to convert the photon energies into equivalent rest mass using E = m*c^2, in order to calculate the mass of the image? And then we should assume one standard Earth gravity in order to convert mass into weight (since weight is a force) ?

[Rasika Devi]

thank you sir, but in matlab there is cmd as graydiffweight() how to use this for image? can you help me?

[Walter Roberson]

I had never heard of that before.

[Rasika Devi]

then how can i complete the weight selection problem for the grayscale image?

[Walter Roberson]

It looks like you just have to pick either a representative pixel or a representative level and call graydiffweight() passing in the data and the location or level.

[Rasika Devi]

Thank you so much sir.. One more thing from that weight value,how do i perform the image segmentation by using any algorithm? kindly help me...

[Walter Roberson]

I suggest you look in the File Exchange, for Image Analyst's Image Segmentation Tutorial

Answer 5

[Syakira Akmal]

May I know how to convert intensity of image inform of histogram?

[Image Analyst]

[counts, grayLevels] = imhist(grayImage);