script for extracting the turbulence around a photo of a building.

turbulence.m

close all
% read the building picture
I1 = imread('IMG_5380.JPG');
% correct the rotation
I2 = imrotate(I1, -90);
% crop it to the area of interest
I3 = imcrop(I2,[378.5 0.5 948 814]);

% focus on an roi to isolate just the sky from the building...
rp = [0 815;70 815;251 435;570 363;949 657;949 0;0 0];
mask = roipoly(I3, rp(:,1)', rp(:,2)');

% histeq just the sky roi...
I4 = roihisteq(I3, rp(:,1)', rp(:,2)');

% then mask the image to remove the building part...
R = I4(:,:,1);
G = I4(:,:,2);
B = I4(:,:,3);
R(~mask) = 0;
G(~mask) = 0;
B(~mask) = 0;
I5 = cat(3,R,G,B);

% convert it to grayscale
BW = rgb2gray(I5);


% jet by itself is too broad to show the structure nicely,  
% so I copy jet across p1 times p2 and truncate to get a 255 size color map 
% then I just played with the value of p1
p1 = 64;
p2 = ceil(256/p1);
cm = repmat(jet(p1),p2,1);
cm = cm(1:255,:);

% construct an image from it
I6 = ind2rgb(gray2ind(BW,255),cm);

% show the final enhancement grayscale with the generated colormap.
imshow(I6);

% and make sure it's in the original aspect ratio.
daspect([1,1,1]);

Uses: ROIHISTEQ, Image Processing Toolbox, MATLAB

The original and enhanced images.