You are on page 1of 7

Lecture 8

Digital Image Processing


Prof. Shih-Fu Chang

Digital Video and Multimedia Lab


Columbia University
Feb. 13th 2006
Most images are downloaded from the web site of the Gonzalez/Woods textbook

„ Last lecture:
„ Image Enhancement: Histogram
Equalization
„ This lecture:
„ Image Enhancement
„ Spatial operator
„ Derivative operators: Laplacian, Unsharp
Masking
„ Readings: G&W Chap 3

Lecture #8 1
Point Operation: Contrast Stretching

Histogram Equalization

„ Note the output


still does not have
ideal uniform
distribution due to
discrete domains

Lecture #8 2
Spatial domain filtering – averaging
multiple frames

Spatial domain filtering – averaging


multiple pixels

Lecture #8 3
Spatial domain filtering – mean vs.
median

average median

Profiles of image intensity

Lecture #8 4
Derivative operations: Laplacian

−∇2 f ( x, y )

f ( x, y ) −∇2 f ( x, y )
−α∇2 f ( x, y )

Sharpening filter: high-boost with


Laplacian

−∇2 f ( x, y )

1.7 f ( x, y ) −∇2 f ( x, y )
f ( x, y ) −∇2 f ( x, y )

10

Lecture #8 5
Detect edge locations: Sobel operator

11

Combination of different operations

Original Laplacian

Sharpened with Sobel „ Good at


Laplacian enhancing
edges
„ Good at
enhancing
detailed
structures

12

Lecture #8 6
Principal rules + trial and error

Smoothed Smooth Sobel x


Laplacian Sharpened
Sobel

Sum of Power Law


Mask & image

13

Lecture #8 7

You might also like