CSE4203: Computer Graphics

Chapter – 8 (part - B)
Graphics Pipeline
Mohammad Imrul Jubair
Outline
• Bresenham’s Circle Drawing Algorithm

2
Assumptions

Given,
Radius R circumference

We have to develop an algorithmthat

generates this circumference

3
Assumptions
The first pixel of the circumference is plotted on (0, R)
Given,
Radius R

(0, R)

(0,0)

4
Observation
The first pixel of the circumference is plotted on (0, R)
Then the plotting of next pixels starts clock-wise….

That means the plotting

(X, Y) = (0, R) 2 starts from (0,R) and
3 moving into the 2nd Octant

4 1

(0,0)
8
5

6 7

5
Observation

while moving through the 2nd octant, theXvalue

is increasing and Y value is decreasing

6
Observation

(0,R)

(R,0)

7
Observation

(0,R) (X,Y)

(R,0)

8
Observation

(0,R) (X,Y)

X=Y

45o
(R,0)

9
Observation

(0,R) (X,Y)

X=Y

45o (Y,X)
(R,0)

10
Observation

(0,R) (X,Y) In 2nd Octant : X<Y

But, in 1 st Octant : X>Y

X=Y

45o (Y,X)
(R,0)

11
Observation

(X,Y)

(-X,Y)
So, if we can obtain (X,Y) in 2nd
octant, we can calculate the
(Y,X) points-
• 7th Octant : (X,-Y)
• 6th Octant : (-X, -Y)
• 3rd Octant : (-X, Y)

(X,-Y)
(-X,-Y)

12
Observation

(X,Y)

(-X,Y)
So, if we can obtain (X,Y) in 2nd
octant, we can simply swap X
(Y,X) and Y to get the points-
(-Y,X) • 1 st Octant : (Y, X)
• 8th Octant : (Y, -X)
• 5th Octant : (-Y, -X)
(-Y,-X) (Y,-X) • 4th Octant : (-Y, X)

(X,-Y)
(-X,-Y)

13
Drawing in all the octants

So, if we can obtain

(X,Y) in 2nd octant, we
can calculate the (0,R)
points in other 7 So, our target is
octants to get the pixels
of only 2nd octant
of the
(0,0)
circumference

14
Bresenham’s Circle Drawing Algorithm: How it works

(0, R) C

15
Bresenham’s Circle Drawing Algorithm: How it works

(0, R) C E Next pixel is chosen

(from E or SE) to build
the line successively

SE

16
Bresenham’s Circle Drawing Algorithm: How it works

(0, R) C E Next pixel is chosen

(from E or SE) to build
the line successively

SE

17
Bresenham’s Circle Drawing Algorithm: How it works

(0, R) C E Next pixel is chosen

(from E or SE) to build
the line successively

SE

18
Bresenham’s Circle Drawing Algorithm: How it works

(0, R) C E Next pixel is chosen

(from E or SE) to build
the line successively

SE

19
Bresenham’s Circle Drawing Algorithm: How it works

(0, R)
Next pixel is chosen
(from E or SE) to build
the line successively
E

SE

20
Bresenham’s Circle Drawing Algorithm: How it works

(0, R)
Next pixel is chosen
(from E or SE) to build
the line successively
E

SE

21
Bresenham’s Circle Drawing Algorithm: How it works

(0, R)
As we know that,
In 2nd Octant : X < Y
In 1st Octant : X > Y

We will stop when X > Y,

C that means when 2nd octant
is completed

22
Equation of Circle and its function representation

(X,Y)

23
Equation of Circle and its function representation

If F(X,Y) = 0, the point (X,Y) on

the circle

If F( X, Y ) > 0, the point (X,Y)

is outside the circle

If F( X, Y ) < 0, the point (X,Y)

is inside the circle

24
Selecting E or SE

C(XP,YP)

25
Selecting E or SE

Selecting E or SE depends
on closeness to the
circumference.
C(XP,YP) E(XP+1,YP)
If E is closer to
circumference,
then E is selected

If SE is closer,
then SE is selected
SE(XP+1,YP-1)

26
Selecting E or SE using Mid Point Criteria

If midpoint M is outside
the circle, SE is closer to
the circumference,
So, SE is selected

If midpoint M is inside
the circle, E is closer to
the circumference,
So, E is selected

27
Selecting E or SE using Mid Point Criteria

We know,
Lets put the mid point M’s coordinate in function F(X,Y)
F(M) = F(XP + 1, YP – 0.5) = (XP + 1)2 + (YP – 0.5)2 – R2

C (XP ,YP) E(XP+1, YP)

Lets store F(M) in a variable d

(XP+1, YP-0.5)M So, d = F(M)

d is called ‘decision variable’

SE(XP+1,YP-1)

28
Selecting E or SE using Mid Point Criteria

If d >= 0, then midpoint M is

outside the circle, SE is closer
to the circumference,
So, SE is selected

If d < 0, then midpoint M is

inside the circle, E is closer
to the circumference,
So, E is selected

29
Bresenham’s Mid Point Criteria : Successive Updating (for selecting E)

d1 =F(M1)
= F(XP+1, YP- 0.5)
= (XP+1)2 + (YP- 0.5)2 - R2
C(XP,YP) E

M1

SE

30
Bresenham’s Mid Point Criteria : Successive Updating (for selecting E)

d1 =F(M1)
= F(XP+1, YP- 0.5)
= (XP+1)2 + (YP- 0.5)2 - R2
C(XP,YP) E
If d1 < 0, E(XP=XP+1, YP)

M1

SE

31
Bresenham’s Mid Point Criteria : Successive Updating (for selecting E)

d1 =F(M1)
= F(XP+1, YP- 0.5)
= (XP+1)2 + (YP- 0.5)2 - R2
C E
If d1 < 0, E(XP=XP+1, YP)

d2 = F(M2)
M2
= F(XP+2, YP-0.5)
= (XP+2)2 + (YP- 0.5)2 - R2
= XP2+ 4XP + 4 + (Y P- 0.5)2 -R2
SE = XP2+ 2XP + 1 + (YP - 0.5)2 - R2 + 2XP + 3
= d1 + (2XP+3)

32
Bresenham’s Mid Point Criteria : Successive Updating (for selecting E)

d1 =F(M1)
= F(XP+1, YP- 0.5)
= (XP+1)2 + (YP- 0.5)2 - R2
C E
If d1 < 0, E(XP=XP+1, YP)

d2 = F(M2)
M2
= F(XP+2, YP-0.5)
= (XP+2)2 + (YP- 0.5)2 - R2
= XP2+ 4XP + 4 + (Y P- 0.5)2 -R2
SE = XP2+ 2XP + 1 + (YP - 0.5)2 - R2 + 2XP + 3
= d1 + (2XP+3)

33
Bresenham’s Mid Point Criteria : Successive Updating (for selecting E)

Every iteration after selecting E, we can

successively update our decision variable with-

dNEW = dOLD+ (2XP + 3)

34
Bresenham’s Mid Point Criteria : Successive Updating (for selecting SE)

d1 =F(M1)
= F(XP+1, YP- 0.5)
= (XP+1)2 + (YP- 0.5)2 - R2

C(XP,YP) E

M1

SE

35
Bresenham’s Mid Point Criteria : Successive Updating (for selecting SE)

d1 =F(M1)
= F(XP+1, YP- 0.5)
= (XP+1)2 + (YP- 0.5)2 - R2

C(XP,YP) If d1 >= 0, SE(XP=XP+1, YP-1)

E

M1

SE

36
Bresenham’s Mid Point Criteria : Successive Updating (for selecting SE)

d1 =F(M1)
= F(XP+1, YP- 0.5)
= (XP+1)2 + (YP- 0.5)2 - R2
If d1 >= 0, SE(XP=XP+1, YP-1)
d2 = F(M2)

C E
…. DIY….

M2
= d1 + (2XP- 2YP+5)
SE

37
Bresenham’s Mid Point Criteria : Successive Updating (for selecting SE)

Every iteration after selecting NE, we can

successively update our decision variable with-

dNEW = dOLD + (2XP- 2YP+5)

38
Bresenham’s Mid Point Criteria : Successive Updating (summary)

If d < 0, then midpoint M is

inside the circle, E is closer
to the circumference,
So, E is selected and do-
d = d + ΔE
Where, ΔE = 2XP + 3

If d >= 0, then midpoint M is

outside the circle, SE is closer
to the circumference,
So, SE is selected and do-
d = d + ΔSE
Where, ΔSE = 2XP - 2YP + 5

39
Initialization

C(0,R) E(XP+1, YP) dINIT= F(M1)

= F(1, R-0.5)
= (1)2 + (R –0.5)2 –R2
(1, R-0.5) M1 = 1 + R2–R + 0.25 –R2
=1.25 –R

SE(XP+1, YP-1)

40
Initialization

C(0,R) E(XP+1, YP) dINIT= F(M1)

= F(1, R-0.5)
= (1)2 + (R –0.5)2 –R2
(1, R-0.5) M1 = 1 + R2–R + 0.25 –R2
=1.25 –R
≈ 1 -R
SE(XP+1, YP-1)

41
Initialization

R=2
d = 1.25 – R = – 0.75

R=2
d = 1 – R = –1

42
Summary

So, finally…..

dINIT = 1 – R

If d < 0, then E is selected, d = d + ΔE

If d >= 0, then SE is selected, d = d + ΔSE

Where,

ΔE = 2XP + 3
ΔSE = 2XP – 2YP + 5

43
 Algorithm

void MidpointCircle(int radius)

{
int x = 0;
int y = radius ;
int d = 1 –radius ;
CirclePoints(x, y);
while (y > x)
{
if (d < 0) /* Select E*/
d = d + 2 * x + 3;
else
{ /* Select SE*/
d = d + 2 * ( x –y ) + 5;
y = y –1;
}
x = x + 1;
CirclePoints(x, y);
}
}
44
 Algorithm

void MidpointCircle(int radius)

{
int x = 0;
int y = radius ;
int d = 1 –radius ; CirclePoints (x,y)
CirclePoints(x, y); Plotpoint(x,y) ;
while (y > x) Plotpoint (x,-y) ;
{ Plotpoint(-x,y) ;
if (d < 0) /* Select E*/ Plotpoint(-x, -y) ;
d = d + 2 * x + 3; Plotpoint(y,x) ;
else Plotpoint(y, -x) ;
{ /* Select SE*/ Plotpoint(-y, x) ;
d = d + 2 * ( x –y ) + 5; Plotpoint( -y, -x) ;
y = y –1; end
}
x = x + 1;
CirclePoints(x, y);
}
}
45
Example

0 1 2 3 4 5 6 7 Given:
10 Radius , R = 10
9
8
7
6
5
4

46
Example

0 1 2 3 4 5 6 7 Given:
10 Radius , R = 10
9 (x,y)=(0,10)
8 h = 1 –R= -9
7
6
5
4

47
Example

0 1 2 3 4 5 6 7 Given:
10 Radius , R = 10
9 (x,y)=(0,10)
8 h = 1 –R= -9
7
6
5
4

K 1
2x 0
2y 20
h
(x,y)

48
Example

0 1 2 3 4 5 6 7 Given:
10 Radius , R = 10
9 (x,y)=(0,10)
8 h = 1 –R= -9
7
6
5
4

K 1
2x 0
2y 20
h
(x,y) E(1,10)
h <= 0,E
49
Example

0 1 2 3 4 5 6 7 Given:
10 Radius , R = 10
9 (x,y)=(0,10)
8 h = 1 –R= -9
7 h = h+ΔE=h+2x+3
6 =-9+0+3
=-6
5
4

K 1
2x 0
2y 20
h -6
(x,y) E(1,10)

50
Example

0 1 2 3 4 5 6 7 Given:
10 Radius , R = 10
9 (x,y)=(0,10)
8 h = 1 –R= -9
7
6
5
4

K 1 2
2x 0 2
2y 20 20
h -6
(x,y) E(1,10)

51
Example

0 1 2 3 4 5 6 7 Given:
10 Radius , R = 10
9 (x,y)=(0,10)
8 h = 1 –R= -9
7
6
5
4

K 1 2
2x 0 2
2y 20 20
h -6
(x,y) E(1,10) E(2,1 0)
h <= 0,E
52
Example

0 1 2 3 4 5 6 7 Given:
10 Radius , R = 10
9 (x,y)=(0,10)
8 h = 1 –R= -9
7 h = h+ΔE=h+2x+3
6 =-6+2+3
=-1
5
4

K 1 2
2x 0 2
2y 20 20
h -6 -1
(x,y) E(1,10) E(2,1 0)

53
Example

0 1 2 3 4 5 6 7 Given:
10 Radius , R = 10
9 (x,y)=(0,10)
8 h = 1 –R= -9
7
6
5
4

K 1 2 3
2x 0 2 4
2y 20 20 20
h -6 -1
(x,y) E(1,10) E(2,1 0)

54
Example

0 1 2 3 4 5 6 7 Given:
10 Radius , R = 10
9 (x,y)=(0,10)
8 h = 1 –R= -9
7
6
5
4

K 1 2 3
2x 0 2 4
2y 20 20 20
h -6 -1
(x,y) E(1,10) E(2,1 0) E(3,1 0)
h <= 0,E
55
Example

0 1 2 3 4 5 6 7 Given:
10 Radius , R = 10
9 (x,y)=(0,10)
8 h = 1 –R= -9
7 h = h+ΔE=h+2x+3
6 =-1+4+3
=6
5
4

K 1 2 3
2x 0 2 4
2y 20 20 20
h -6 -1 6
(x,y) E(1,10) E(2,1 0) E(3,1 0)

56
Example

0 1 2 3 4 5 6 7 Given:
10 Radius , R = 10
9 (x,y)=(0,10)
8 h = 1 –R= -9
7
6
5
4

K 1 2 3 4
2x 0 2 4 6
2y 20 20 20 20
h -6 -1 6
(x,y) E(1,10) E(2,1 0) E(3,1 0) S(4,9)
h > 0,SE
57
Example

0 1 2 3 4 5 6 7 Given:
10 Radius , R = 10
9 (x,y)=(0,10)
8 h = 1 –R= -9
7 h = h+ΔSE=h+2x-2y+5
6 =6+6-20+5
=-3
5
4

K 1 2 3 4
2x 0 2 4 6
2y 20 20 20 20
h -6 -1 6 -3
(x,y) E(1,10) E(2,1 0) E(3,1 0) S(4,9)

58
Example

0 1 2 3 4 5 6 7 Given:
10 Radius , R = 10
9 (x,y)=(0,10)
8 h = 1 –R= -9
7
6
5
4

K 1 2 3 4 5 6 7
2x 0 2 4 6 8 10 12
2y 20 20 20 20 18 18 16
h -6 -1 6 -3 8 5 6
(x, y) E(1,10) E(2,10) E(3,10) S(4,9) E(5,9) S(6,8) S(7,7)

59
Example

0 1 2 3 4 5 6 7 Given:
10 Radius , R = 10
9 (x,y)=(0,10)
8 h = 1 –R= -9
7
6
5
4
Untill y >x
K 1 2 3 4 5 6 7
2x 0 2 4 6 8 10 12
2y 20 20 20 20 18 18 16
h -6 -1 6 -3 8 5 6
(x, y) E(1,10) E(2,10) E(3,10) S(4,9) E(5,9) S(6,8) S(7,7)

60
Practice Problem
• Perform the midpoint algorithm to draw a circle’s portion at 7th octant
which has center at (2,-3) and a radius of 7 pixels. Show each
iterations and plot the points.

61