counting pixel of an image?

i have used the thresholding for image ..

the command is

length(find(abs(cw)<=50));
im(length(abs(cw)<=50))=0 ;

(1)for examble 42447 coeffients made to be zero out of 65656. my doubt is whether that remaing (23089 ) pixel only present in that image or else 0 will be presented for the whole 65536 ..

(2)how to count or calculate the number of pixels present in that image ?

(3)i apply the wavelet transform.. after that i apply this can i use this threshold for compression ?

pleased to ask reply from all ..

[vishnu]

length(find(abs(cw)<=50));

cw(length(abs(cw)<=50))=0 ;

sir ,

(1)length is i calculated to know the how many pixels are present under 50 value

(2) can you explain what is the difference between pixel and coefficients and pixels

(3)after wavelet transform some pixels are made to zero with above code can i say this as compression ...

Answer 1

[Image Analyst]

Responding to your comment

(1) The number of pixels under 50 is given more simply by

numPixelsLessThan50 = sum(sum(cw<50));

No need for find() and abs().

(2) A pixel contains the value of your image array at a certain location. I don't know what you are calling coefficients. I don't have the wavelet toolbox so I can't answer any questions about that.

(3)I don't understand what #3 is saying or asking. Something about wavelets, setting to zero, and compression but it's not clear. Ask someone else to reword it.

[vishnu]

thank you very much

[Nivetha m]

i want to be segmenting below image using k means algorithm and also i find after segmentation how many cells in the image

[Walter Roberson]

please create a new Question for that.

[shivam sahil]

Your code is specific for grayscale Image, I want to count number of rgb codes for example r=223 g=92 b=75, how many pixels in image are with that composition. If that is the case what code should I use?

[Image Analyst]

I'm attaching my RGB program for kmeans(). Generally I feel that kmeans is not a good method for segmentation of RGB images, but for whatever it's worth, it's attached. Don't say I didn't warn you if you don't like the results.

Example based discriminant analysis is better and I'm attaching a demo for that that uses the classify() function. There may be even better methods.

[Madhava Teja Munagala]

Take input image, Then based on threshold Img divide into foreground and background, After find means for forground,and background,

      if true
        % code
      end
After that compare pixel with means of both fg ang bg ,
These is concept,how to write program by using otsu 2d

Answer 2

[Image Analyst]

Why is length() in there? What are the coefficients? Do you mean pixels instead of coefficients? What is the difference between cw and im?

The number of pixels in an image is:

[rows columns numberOfColorChannels] = size(im);
numberOfPixels = rows*columns;

or

numberOfPixels = rows*columns*numberOfColorChannels;

depending on how you're defining pixel.