How can i split 2 type of objects ( protrusions and depressions) in 2 different images ?

            if true
             close all ; 
clear all;
clc ;

%% read the image i = imread('a.jpg');

%% convert to grayscale x1 = rgb2gray(i);

%% preprocessing % adjusment

ww = fspecial('gaussian' ,[10 10 ] , .66 ); % lowpass gaussian filter www= imfilter(x1,ww,'conv');

x=imadjust(www , [] , [] ,5);

kk = 255-x; %complement figure(2), imshow(kk);

%Histogram of image figure(1); [pixelCount, grayLevels] = imhist(kk); bar(pixelCount); title('Histogram of original image');

%median filter A = medfilt2(kk,[2 2]); J = stdfilt(A);

% fill a gap in the pen's cap se = strel('disk',3); bw = imopen(kk,se);

% fill any holes, so that regionprops can be used to estimate % the area enclosed by each of the boundaries bw = imfill(bw,'holes');

%% centroids

xy= im2bw(kk); bw = imfill(xy,'holes');

se = strel('disk',5); xy=imdilate(xy,se); grayImage = xy ;

[l , num] = bwlabel(bw); stat = regionprops(l,'centroid'); %%%%%%%%%%%STATS = regionprops(bw, properties) ; figure(8),imshow(bw); hold on; for x = 1: num plot(stat(x).Centroid(1),stat(x).Centroid(2),'ro'); end

%%%segmenting cc = bwconncomp(bw); stats = regionprops(cc, 'Area','Eccentricity'); idx = find([stats.Area] < 15 ); idx2 = find([stats.Area] > 15 );

BW1 = ismember(labelmatrix(cc), idx); BW2 = ismember(labelmatrix(cc), idx2); figure(5),imshow(BW1);

figure(6),imshow(BW2);

%%% bad results

            endHello
I need help in segmentation ... I have a double sided Braille image I wont to segment the protrusions and the depressions in 2 different images .
I'm lost .........

I converted the image to binary when trying to split the 2 types of objects into 2 images ... I didn't know how to do it ?????

how could I recognise 2 type of objects ?

[Adam]

Please format your code to make it readable for anyone who might try to answer your question.