Try this:
D = readmatrix('Boundary_closed_1s_3s.txt');
figure
[c,h] = contour(D, [1 1]*0.5);
Levels = h.LevelList;
idx = find(c(1,:) == Levels(1));
Len = c(2,idx);
for k = 1:numel(idx)
x{k} = c(1,idx+1:Len(k));
y{k} = c(2,idx+1:Len(k));
end
[b,v] = boundary(x{1}(:), y{1}(:)); % Introduced In R2014b
figure
fill(x{1}(b), y{1}(b), 'r')
title('Boundary')
figure
fill(x{1}, y{1}, 'g')
title('Filled Contour')
It uses the contour function to extract the outlines of the triangle, then first boundary (R2014b and later) to fill the triangle, and then a second one to fill the contour. Choose the one you want.
For figure(1):
.
EDIT — (5 Feb 2021 at 03:06)
The ‘v’ output of boundary is the area it encloses. I assume it corresponds to pixels, however I cannot determine that. Nothing in the original Question mentioned anything about pixels. It might be possible to use ‘x{1}(b)’ and ‘y{1}(b)’ with the original data to outline tha area in the image. The ‘y’ axis direction in the image will be reversed from the plot direction.
