please help me with the problem for dividing an image into blocks.i have given the algorithm.

I am working on super pixel segmentation algorithm for detecting cloud images form clear sky images based on pixel intensity values.in this super pixel segmentation we want to divide an image into irregular blocks base on texture similarity, brightness similarity and contour continuity for detecting cloud images from clear sky images.here are the images I can give for super pixel segmentation.

figure c is division of an image based on super pixel segmentation.

I can give you the algorithm for super pixel segmentation.

Input: an RGB cloud image, a division of cloud image after applying superpixel segmentation.

Output: detection result.

step1: transfer RGB image to R-B feature image, which is simply difference of the R-channel and B-channel;

step2: normalize the elements of R-B feature image to the range of 0 ∼ 255;

step3: calculate local threshold for each superpixel on the R-B feature image;

step4: calculate the threshold matrix by bilinear interpolation using the local thresholds of all of the superpixels;

step5: compare the threshold matrix with R-B feature image pixel by pixel to get the final detection result.

if threshold is >0 it is cloud otherwise clear sky.

can u plz help me...I provided a link for the paper

[Geoff Hayes]

Shamili - what in particular do you need help with? If you have started to implement the algorithm in MATLAB, then what problems are you experiencing? Please try to be specific and include any error messages that you are observing.

Answer 1

[Image Analyst]

I'm not sure what is wrong with your code, if anything. You did not attach it or the paper. Anyway, the FAQ would probably apply : http://matlab.wikia.com/wiki/FAQ#Can_you_program_up_the_algorithm_in_this_article_for_me_and_explain_it_to_me.3F

Here is a similar algorithm: http://www.mathworks.com/matlabcentral/fileexchange/37197-dem--diffused-expectation-maximisation-for-image-segmentation - maybe you can just use that.