11/24/2024

Objectives

Lecture No. 7
• Image interpolation

Digital Image Fundamentals • Relationships between pixels

• Distance measures

CIS-322

Image Interpolation Image Interpolation (contd.)

• The process of using known data to estimate values at • Three techniques are:
unknown locations
• Image Interpolation is used extensively for  Nearest Neighbor Interpolation
 Zooming
 Shrinking  Bilinear Interpolation
 Rotating
 Geometric corrections  Bicubic Interpolation
• Zooming requires:
 The creation of new pixel locations
 The assignment of gray levels to those new locations

Nearest Neighbor Interpolation (contd.) Nearest Neighbor Interpolation (contd.)

• 500 x 500 sized image to be scaled to 1.5 times

• 750 x 750 grid is prepared with the same pels
spacing as in the above image
• We now compress the grid of 7502 to 5002
• The pixels spacing in the new image is shrunken
• We have to assign the correct intensity to the overlay
• By selecting the proper intensity from the underlay
image, the overlay is ready
• 7502 image is expanded to the original size to get the
actual picture quality

1
 11/24/2024

Nearest Neighbor Interpolation (contd.)

• Method is quite simple:

 Assign to each new location the intensity of its Some Basic Relationships
nearest neighbor between Pixels
• Quick as it is very simple (advantage)

• Hardly used because of straight edges distortion artifact

(disadvantage)

Neighbors of a Pixel Neighbors of a Pixel (contd.)

• 4-neighbors of “p” are the four horizontal and vertical • The four diagonal neighbors of p(x, y) are ND(p)
neighbors of a point p(x, y) as given by N4(p) (x+1, y+1), (x+1, y-1), (x-1, y+1), (x-1, y-1)
(x+1, y), (x-1, y), (x, y +1), (x, y-1)

(x) (x+1)
(x-1) (x) (x+1)
f(x,y) is a yellow circle
(x-1)
f(x,y) is a yellow circle
(y-1) f(x,y-1) is top one (y-1) f(x+1,y-1) is top-right one
f(x-1,y) is left one (y)
f(x-1,y-1) is top-left one
(y)
f(x,y+1) is bottom one f(x-1,y+1) is bottom-left one
(y+1)
(y+1)
f(x+1,y) is right one f(x+1,y+1) is bottom-right one

Neighbors of a Pixel (contd.) Adjacency

• The 4-neighbors together with diagonal neighbors of (x, y) • V is the set of gray-level values used to define
are called 8-neighbors N8(p) of (x, y)
adjacency (e.g. V={1} for adjacency of pixels of
value 1)
• Three types are:
(x) (x+1)
(x-1)
f(x,y) is a yellow circle
(y-1) (x-1,y-1), (x,y-1),(x+1,y-1),  4-Adjacency
(y)
(x-1,y), (x,y), (x+1,y),  8-Adjacency
(x-1,y+1),(x,y+1), (x+1,y+1)  m-adjacency
(y+1)

2
 11/24/2024

Adjacency (contd.) Adjacency (contd.)

• 4-Adjacency: Two pixels p and q with values from V • 8-Adjacency: Two pixels p and q with values from
are 4-adjacent if q is in the set N4(p), V ={1} V are 8-adjacent if q is in the set N8(p), V ={1}

0 1 1 0 0 1 1 0

1 1 0 1 1 1 0 1

0 0 1 1 0 0 1 1

Adjacency (contd.) Adjacency (contd.)

• m-adjacency (mixed adjacency): Two pixels p • m-adjacency (mixed adjacency)

and q with values from V are m-adjacent if
 q is in N4(p), or
0 1 1 0
 q is in ND(p) and the set N4(p)∩N4(q) has no
pixels whose values are from V
1 1 0 1
• Mixed adjacency is a modification of 8-adjacency
and is used to eliminate the multiple path
connections that often arise when 8-adjacency is 0 0 1 1

used

Adjacency (contd.) Path

• A path from pixel p(x, y) to pixel q(s, t) is a sequence of
pixels with coordinates (x0, y0),(x1, y1)………(xn, yn) such
that
 (x0, y0) = (x, y)
 (xn, yn) = (s, t)
 Intermediate pixels are adjacent (n is the length of the path)

• For a closed path (xo, yo) = (xn, yn)

• 4-, 8- and m-paths may be defined depending on the type
of adjacency specified

3
 11/24/2024

Path Connectivity

• S = Subset of pixels in an image

• Two pixels p and q are connected if path is
between them in S (lying in V)
• For any pixel p in S, the set of pixels that are
connected to it in S is called a connected
8-paths m-paths component of S
• If S has only one connected component then S is
called a connected set

Region Region (contd.)

• R = Subset of pixels in an image. We call R a region • Only 4- and 8- adjacency is considered for regions
of the image if R is a connected set • Ri and Rj are considered adjacent by 8-adjacency
only (no 4 path exists between the two regions, so their
• Regions not connected are disjoint union is not a connected set in that case)

• Ri and Rj are said to be adjacent if their union forms

a connected set

Foreground and Background of Image Boundary

• The boundary of a region R is constituted of the set of
points that are adjacent to points in the complement of R
• An image have K disjoint regions Rk, where k = 1,
2, 3,…, K • In another way, it is set of pels that have at least one
background neighbor
• Ru = Union of all the K regions • Regional points and background points are defined in
terms of 8-connectivity (4-connectivity fails between
• (Ru)c = Complement of Ru background and region)

• All points in Ru are called the Foreground of the

image
• All points in (Ru)c denoted as Background of the
image

4
 11/24/2024

Edge Distance Measures

• The boundary of a finite region forms a closed path
(a global concept) • For pixels p, q, z with coordinates (x, y), (s, t), (u,
• Edges are formed from pixels with derivative v), D is a distance function if:
values that exceed a preset threshold (a local
concept)  D(p, q) ≥ 0 (D(p, q) = 0 iff p = q)
• Edges are based on intensity level measurement at  D(p, q) = D(q, p)
a discontinuity  D(p, z) ≤ D(p, q) + D(q, z)
• Exception: edge and boundary are same in case of
binary images (the edge extracted from binary region
would be same as the region boundary)

Distance Measures (contd.) Distance Measures (contd.)

• The Euclidean distance between p and q is • The D4 distance (also called city-block distance) between p and q
is
𝐷4 𝑝, 𝑞 = 𝑥 − 𝑠 + 𝑦 − 𝑡
De(p, q)=[(x-s)2+(y-t)2]1/2  the pels having a D4 distance from (x, y) less than or equal to
some value r form a diamond centered at (x, y)
[Fig: pixels with D4 ≤ 2 from (x, y) ]
• Pixels having a distance from (x, y) less than or 2
equal to a disc of radius r centered at (x, y) 2 1 2
2 1 0 1 2
2 1 2
2

Distance Measures (contd.) Distance Measures (contd.)

• The D8 distance (or chessboard distance) between p and q is • For example assume p, p2, p4 = 1
𝐷8 𝑝, 𝑞 = max( 𝑥 − 𝑠 , 𝑦 − 𝑡 ) p1, p3 = 0 or 1
 the pels with a D8 distance from (x, y) less than or equal to some
value r form a square centered at (x, y)  If only connectivity of pixels valued 1 is
[Fig: pixels with D8≤2 from (x, y) ] p3 p4 allowed, and p1 and p3 are 0, the m-distance
p1 p2 between p and p4 is 2
2 2 2 2 2
p  If either p1 or p3 is 1, the distance is 3
2 1 1 1 2
2 1 0 1 2
 If both p1 and p3 are 1, the distance is 4
2 1 1 1 2
2 2 2 2 2