Dear Nor,

Secondly, Is it one raindrop or all of them that you want to estimate? I have found the area of a weld before by first turning the picture into black and white (specifically, make the area you want to estimate white) then you apply the command bwarea(X), where the X is the matrix of the picture. That will give you the are a of the on-pixels you can then convert between pixels and m^2

Dear Nor,

My doubt is are you calculating area for the whole image or the area where the rain drop density is high.

In matlab bwarea  function is used to calculate the area .

 var = bwarea(BW)

 i have attach some link realted to u r topic plz find those. 

http://radio.feld.cvut.cz/matlab/toolbox/images/binary12.html

http://in.mathworks.com/help/images/ref/regionprops.html

Even u can use Connected component labeling.after converting image in to binary.

This will give u different white patches in an image and no of pixels interconnected.

So take out each rain drop by calling with different label nos and the use bwarea.

Hi Nor,

If I understand your question correctly, it looks like you just want to find the sum of all the detected raindrops. Is that correct?

In your code, since you calculate the area of each object in the "for" loop at the end (using the statement "area = stats(k).Area;"). You simply need to use an accumulator variable to sum these areas.

Basically, your code requires a slight modification (italicized in the following snippet) -

...

threshold = 0.94;

totalArea = 0;

for k = 1:length(B)

         ...

         area = stats(k).Area;

         totalArea = totalArea + area;

         ...

end

At the end of the loop run, totalArea would contain the sum of the areas of all the detected objects. This is merely the total number of pixels in the detected objects (since regionprops() counts the number of pixels in each object when you specify 'Area' as the property).

Another option is to find the number of white pixels in the final mask (the variable fill in your code) directly. This can be done in a single line -

totalArea = numel(find(fill));

This method should give you the same value as that calculated by accumulation above.

Keep in mind that you can use bwarea() as well (again, on the variable fill), but this will not exactly be equal to the sum of the individual areas calculated in the loop, since bwarea() weighs different patterns of pixels differently.

Try to see regionprop command in MATLAB, it is also better one