Image Enhancement in the

1
Chapter 3 Image Enhancement in the Spatial Domain
http//www.prenhall.com/gonzalezwoods
2
Histogram for a given gray level image 1.
Construct an array H0..255 such that Hk
of pixels in the image that their gray level is
k. ( scan every pixel that if pixel pi,j has
gray level k, then Hk Hk1.) 2. How to
show such histogram? 3. How to find the gray
level that corresponds to the kth percentile of
the histogram?
3
Chapter 3 Image Enhancement in the Spatial Domain
4

• Some Transformations on images
• r original gray value,
• s transformed gray value,
• gray values 0 ..L-1
• Images Negatives
• s (L-1) r

5
Chapter 3 Image Enhancement in the Spatial Domain

• Log Transformations s c log (1 r)

6
Chapter 3 Image Enhancement in the Spatial Domain
7
Chapter 3 Image Enhancement in the Spatial Domain

• Power-Law Transformations s c r?

8
(No Transcript)
9
(No Transcript)
10

• Piecewise-Linear Transformations
• s ( s2 - s1)/(r2 - r1) (r - r1) s1
• Binarization
• Let T(r) be the threshold such that
• s 0 if r lt T(r) and s 255 if r
  gt T(r) .

11
(No Transcript)
12

• Highlights Range A, B of gray levels

13

• Image Subtraction g(x,y) f(x,y) h(x,y).

14

• When we do subtraction of two images the range of
  values is 255 255, thus some sort of scaling
  is required to display the results
• Let min the minimum value in the difference
  image
• Add min to the values of all pixels
• Let max the maximum value in the modified
  image
• Scale all the pixels to interval 0,255 by
  multiplying each pixel by 255/max.

15

• Image Averaging

16
(No Transcript)
17

• Project for this topic
• For a given gray level image
• Produce a histogram.
• Find the gray values for min( gmin), max(
  gmax), and the kth percentile( gk) of the
  histogram.
• Do a piecewise linear transformation such that
  gmin, gk maps into 0, t, gk, gmax maps
  intot1, 255.
• Select two different thresholds, th1 th2, and
  compare the binarised results.