lecture05

lecture05 - ELEC317 Lecture 5 Image Enhancement Digital...

Info iconThis preview shows pages 1–6. Sign up to view the full content.

View Full Document Right Arrow Icon
ELEC317 Digital Image Processing Lecture 5 Image Enhancement Image Enhancement -- This is sharpening of desirable features in image. -- Information inside data is actually decreased. -- Whether “enhanced” image is good usually is subjective to viewer Î No universal quantifying criteria Î Procedures are normally “heuristics” based on empirical observations. Common Image Enhancement Methods: 1. Point Operations -- Operate on individual pixel values * Contrast Stretching * Noise Clipping * Window Slicing * Histogram Modeling / Equalization 2. Spatial Operations -- Often Filtering * Noise Smoothing * Median Filter * Lowpass, Highpass, Bandpass Filtering * Zooming – Enlarge / Reduce size of image * Unsharp Masking 3. Pseudocoloring * False Coloring * Pseudocoloring 1
Background image of page 1

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
2 4. Transform Operations * Linear Filtering via Fast Fourier Transform ( FFT ) * Root Filtering * Homomorphic Filtering A. Point Operations This is a memoryless operation. Each pixel’s gray level ] , 0 [ L u is mapped into a new level ] , 0 [ L v where v = f ( u ) A.1. Contrast Stretching Aims at increasing dynamic range of image. L u b b u a a u v b u v a u u v b a < < < + + = , , 0 , ) ( ) ( γ β α -- The region where slope > 1 is stretched. -- Stretched region is normally region where most pixel values fall in. -- Refer to fig. 7.3 p236. a v a = a b v v a b = b L v L b = L = 255 L L γ β α a b v u b v a v This region stretched. L b v a v a b Histogram L. C. image
Background image of page 2
A.2. Clipping and Thresholding v Clipping u 0 a b L Choosing = = 0 γ α , we have clipping. v b L u b a u b u a v a u v b < = , , , 0 ) ( β -- It is useful when we know that ] , [ b a u Choosing 0 = = and a = b , we have thresholding. v Thresholding u 0 a v b < < = L u a a u v v b , 0 , 0 -- This is good for binary images. -- See figure 7.6. A.3. Digital Negative V = L – u A.4. Intensity Level Slicing v Without background: L = otherwise b u a L v , , 0 u 0 ab 3
Background image of page 3

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
v With background: = otherwise b u a u L v , , Good for extracting information contained in a fixed gray level. u 0 abL L A.5. Bit Extraction = = otherwise bit t significan n if L v th , 1 , 0 n B n u Int i 2 MSB LSB u = k 1 k 2 k 3 k 4 k 5 k 6 k 7 k 8 k n = i n -2i n-1 i 0 = 0 i 1 = k 1 i 2 = k 1 k 2 i 3 = k 1 k 2 k 3 Binary subtraction B = 8 k 2 = i 2 -2i 1 = k 1 k 2 -k 1 0 = k 2 A.6. Range Compression v = c log 10 (1 + |u| ) -- Good when dynamic range too large that small pixels cannot be seen. A.7. Change Detection By subtraction two images, their difference can be seen better. B. Histogram Modeling Histogram – Relative frequency of occurrence of gray levels. hist ( A ) Histogram h (7) h (2) h (0) u 0 1 2 3 4 5 6 7 4
Background image of page 4
B.1. Equalization u F u ( u ) v The idea is to find F u ( u ) such that v = F u ( u ) has a flat histogram.
Background image of page 5

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Image of page 6
This is the end of the preview. Sign up to access the rest of the document.

Page1 / 17

lecture05 - ELEC317 Lecture 5 Image Enhancement Digital...

This preview shows document pages 1 - 6. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online