Contrast stretching transformation in Image Processing

14 views (last 30 days)
I tried to solve for question but i didn't do.
I tried this code for solving.
p1=[0,0];
p2=[75,5];
p3=[140,250];
p4=[255,255];
m1=(p1(1,2)-p2(1,2))/(p1(1,1)-p2(1,1));
m2=(p2(1,2)-p3(1,2))/(p2(1,1)-p3(1,1));
m3=(p3(1,2)-p4(1,2))/(p3(1,1)-p4(1,1));
c1=p1(1,2)-m1*p1(1,1);
c2=p2(1,2)-m2*p2(1,1);
c3=p3(1,2)-m3*p3(1,1);
%% Transformation function
t=[];
for x=0:255
if(x<=p2(1,1))
t=[t (m1*x+c1)];
end
if(x>p2(1,1) && x<=p3(1,1))
t=[t (m2*x+c2)];
end
if(x>p3(1,1) && x<=p4(1,1))
t=[t (m3*x+c3)];
end
end
for n=1:s(1,1)
for m=1:s(1,2)
ot(n,m)=t(a(n,m)+1);
end
end
I don't know how to find answer of this question. I will be glad if you help.
  1 Comment
Rooter Boy
Rooter Boy on 15 Nov 2020
I found answer is 130 but i'm not sure.
I tried this:
r = input("Please input the intensity value, r: ");
fprintf("\n");
r1 = 70;
r2 = 150;
s1 = 30;
s2 = 230;
Point_1 = [0 0];
Point_2 = [r1 s1];
Point_3 = [r2 s2];
Point_4 = [255 255];
x = 1;
y = 2;
%Slope of region 1%
Rise = Point_2(y) - Point_1(y);
Run = Point_2(x) - Point_1(x);
Slope_1 = Rise/Run;
%Slope of region 2%
Rise = Point_3(y) - Point_2(y);
Run = Point_3(x) - Point_2(x);
Slope_2 = Rise/Run;
%Slope of region 3%
Rise = Point_4(y) - Point_3(y);
Run = Point_4(x) - Point_3(x);
Slope_3 = Rise/Run;
if r <= r1
fprintf("Region 1\n");
s = Slope_1*r;
fprintf("The output intensity, s = T(r) is: %.2f\n",s);
elseif (r1 < r) && (r < r2)
fprintf("Region 2\n");
Constant_Offset = s1;
Relative_Run = (r - r1);
s = Slope_2*Relative_Run + Constant_Offset;
fprintf("The output intensity, s = T(r) is: %.2f\n",s);
elseif r >= r2
fprintf("Region 3\n");
Constant_Offset = s2;
Relative_Run = (r - r2);
s = Slope_3*Relative_Run + Constant_Offset;
fprintf("The output intensity, s = T(r) is: %.2f\n",s);
end

Sign in to comment.

Accepted Answer

Image Analyst
Image Analyst on 16 Nov 2020
Wow. Try this much easier way:
r = [70, 150]
s = [30, 230];
output = interp1(r, s, 110) % output is 130.
  4 Comments
Image Analyst
Image Analyst on 17 Nov 2020
If this solved the question, can you "Accept this answer" unless you want to wait for possible other, better answers.
Rooter Boy
Rooter Boy on 17 Nov 2020
Excuse me,
Sir, i accepted your answer.
I sometimes forget it.

Sign in to comment.

More Answers (0)

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!