Separating a group of higher intensity pixels in an image

I have an binary image in which only a particular group of pixels are my area of interest. I want to separate these higher intensity group of pixels in a new figure of want to differentiate them with a different color. How can I do it? Please have a look at the link for clarification

https://picasaweb.google.com/lh/photo/qV1R3vavQQurPOPquG8WmpOzMXS1UfqXPDW9N1lFu0w?feat=directlink

Answer 1

[Sean de Wolski]

I would do an:

cirx = imfill(bw, 'holes');

on your circles. Then do a logical AND on this to get the areas inside the circles that are also white. Then maybe bwlabel to color them. I.e:

both = BW&cirx;
L = bwlabel(both);
imshow(label2rgb(L));

[Niranjan]

No Sean de, I am sorry If I have mislead you. The circles were drawn by me for identification of the regions. The original image does not contain these circles. I just want to separate out the aggregated pixels in an image eg) the circled ones. Hope I didn't confuse you this time :)

Answer 2

[Sean de Wolski]

Then use:

bwareaopen

to remove small blobs.

[Niranjan]

I AM GETTING THIS ERROR WHEN I USE bwareaopen(BW,P), CAN YOU TELL ME WHAT SHOULD BE WRONG??

----------------------------------------------------------

??? Error using ==> imageDisplayParseInputs>validateCData at 291

If input is logical (binary), it must be two-dimensional.

Error in ==> imageDisplayParseInputs at 207

common_args.CData = validateCData(common_args.CData,image_type);

Error in ==> imshow at 199

[common_args,specific_args] = ...

Error in ==> mytry at 77

figure, imshow(bwAreaOpenBW);

>>

[Niranjan]

This is the code I have used

bwAreaOpenBW = bwareaopen(I2I6,50);

figure, imshow(bwAreaOpenBW);

[Sean de Wolski]

what is the result when you type:

size(BW);

is it [something, something, 3]?

[Niranjan]

[ 900 1200 3]

[Niranjan]

Can you tell me where the problem is ?

[Sean de Wolski]

So you have all 3-dimensions of your rgb image binarized.

The easiest way to view it would just be to do:

imshow(uint8(bwAreaOpenBW)*255);

But I get the feeling that you're not understanding everything going on under the hood. Do you want a convert your rgb image to gray scale and then do the logical/morphological operations on them? Do you want to binarize all three dimensions (as you have) and then come up with some criteria based on that e.g. if true in two planes leave true, else no. Etc....

Answer 3

[pratibha rane]

Code: BW = imread('E:\ProjectCode\45.jpg');%Binary image BW2 = bwareaopen(BW, 24); figure,imshow(BW2)

Error: Error using ==> imageDisplayParseInputs>validateCData at 291 If input is logical (binary), it must be two-dimensional.

I want to highlight only object area with white color