how to do windowing in image using matlab?

how to do windowing in image using matlab?

[SRI]

Hi shah Could you please clear with your question so that we can assist you with the solution

[Sheema Khattak]

Hello!

I am talking about the window filter in image which works portion by portion in an image.Divide the pixels of an image into frames or blocks to work on it and these blocks are made on the basis of similarity of pixels and its neighbors.thank you

Answer 1

[Image Analyst]

Try conv2(), imfilter(), or nlfilter(), or blockproc() depending on exactly what you want to do. conv2() and imfilter() are for linear filters. nlfilter is for any bizarre custom function you want write and apply to each window. blockproc() is if you want to process windows in "jumps" where it moves from one location to the next by more than 1 pixel.

[Image Analyst]

shah, what's the status? Do you have it working? If not, post your image and the kind of filter you want to do, otherwise, mark the Answer as accepted.

[Sheema Khattak]

What about fwind1(),fwind2(),fspecial(),fspace()etc?Aren't they relevant?whats the difference among functions like fwind1() and nlfilter()?

[Image Analyst]

I never use fwind1() or fwind2() - you don't need them to do linear filters since you can use conv2() and nlfilter() instead. conv2 is convolution which flips the kernel as convolution required. imfilter() just keeps the kernel as it is (doesn't flip) and multiplies and sums. Same as convolution except without the flip. Makes no different for symmetrical filters. fspecial is used to build kernels (filters) of certain patterns, like Gaussians. nlfilter() is used to build non-linear filters - something other than multiply and sum - which conv2() and imfilter() can't do.

[Sheema Khattak]

Would conv2() and imfilter() work on the basis of similarity of pixels and group the similar ones to form a clump of object or not?

[Image Analyst]

No. To do that you need to do connected components labeling. But even before that you need to do image segmentation to find the "objects". See my Image Segmentation Tutorial: http://www.mathworks.com/matlabcentral/fileexchange/?term=authorid%3A31862

Answer 2

[SRI]

clc;

close all;

input = imread('cameraman.tif');

[row,col] = size(input);

% minValue = mean(mean(input));

for i = 2:row-1

    for j =2:col-1

        outImage(i-1:i+1,j-1:j+1) = input(i-1:i+1,j-1:j+1)

% if outImage(i-1:i+1,j-1:j+1) < minValue % outImage(i-1:i+1,j-1:j+1) = 1;

% end

    end
end

Hi shah you just apply any filter operation with this code because this will divide a image into 3 *3 matrix after that you can process as you wish