Title: 60520 Presentation Image Filters
160-520 PresentationImage Filters
School of Computer Science University of
Windsor November 2003
2Outline
- Introduction
- Spatial Filtering
- Smoothing
- Sharpening
- Frequency-Domain Filtering
- Low pass
- High pass
- Summary
3Introduction
- Filtering is the process of replacing a pixel
with a value based on some operations or
functions. - The operations/functions used on the original
image are called filters. - or masks, kernels, templates, windows
4Introduction
- In digital image processing, filters are usually
used to - suppress the high frequencies in an image
- i.e., smoothing the image
- suppress the low frequencies in an image
- i.e., enhancing or detecting edges in the image
5Introduction
- Image filters fall into two categories
- Spatial domain
- Filters are based on direct manipulation of
pixels on an image plane. - Frequency domain
- Filters are based on modifying the Fourier
transform (FT) of an image.
6Spatial Filters
- The general processes can be denoted by the
expression - f(x,y) is the input image
- g(x,y) is the processed image
- T is an operator on f, defined over some
neighborhood of (x,y)
7Spatial Filters
- The principal approach in defining a neighborhood
about a point (x,y) - use a subimage area centered at (x,y)
- shapes of the neighborhood
- circle
- square
- rectangular
8Spatial Filters
Example 33 neighborhood about a point (x,y) in
an image
9(No Transcript)
10Spatial Filters linear filters
- For linear spatial filtering, the result, R, at a
point (x,y) is -
- Rw(-1,-1)f(x-1,y-1) w(0,-1)f(x,y-1)
- w(0,0)f(x,y)
- w(0,1)f(x,y1) w(1,1)f(x1,y1)
11Spatial Filters convolution
- In general, linear filtering of an image is given
by the expression - The image f is of size MN
- The filter mask is of size mn
- m2a1, n2b1
12Spatial Filters smoothing
- Smoothing filters are used for blurring and for
noise reduction. - Smoothing, linear spatial filter
- average filters
- reduce sharp transitions
- side effect
13Spatial Filters smoothing, linear
Gaussian noise
Original
33 mean filter
55 mean filter
14Spatial Filters smoothing, linear
Salt and pepper
33 mean filter
55 mean filter
15Spatial Filters smoothing, linear
- Weighted average filters
- example
- general expression
16Spatial Filters smoothing, nonlinear
- Order-statistic filters
- nonlinear spatial filters
- order/rank the pixels contained in the image area
encompassed by the filter
17Spatial Filters smoothing, nonlinear
- Median filters
- replace a pixel value with the median of its
neighboring pixel values - example
Neighborhood values 15, 19, 20, 23, 24, 25, 26,
27, 50 Median value 24
18Spatial Filters smoothing, nonlinear
- Median filters
- have excellent noise-reduction capabilities
V.S.
Gaussian noise removed By 33 median filter
Gaussian noise removed by 33 mean filter
19Spatial Filters smoothing, nonlinear
- Median filters
- are particularly effective in salt pepper
V.S.
Salt pepper removed By 33 median filter
Salt pepper removed by 33 mean filter
20Spatial Filters smoothing, nonlinear
- Max filters
- maximum of neighboring pixel values
- useful for finding the brightest points in an
image - Min filters
- minimum of neighboring pixel values
- useful for finding the darkest points in an image
21Spatial Filters sharpening
- Principal objective
- highlight fine detail in an image
- enhance detail that has been blurred
- Sharpening can be accomplished by spatial
differentiation
22Spatial Filters sharpening
- For one dimensional function f(x)
- first order derivative
- second order derivative
23Spatial Filters sharpening
- A sample
- (a) a scan line (b) image strip
- (c) first derivative (d) second derivative
(a) (b) (c) (d)
24Spatial Filters sharpening
- The Laplacian
- second derivative of a two dimensional function
f(x,y) - f(x1,y)f(x-1,y)f(x,y1)f(x,y-1)
- -4f(x,y)
25Spatial Filters sharpening
- The Laplacian
- use a convolution mask to approximate
26Spatial Filters sharpening
27Spatial Filters sharpening
28Frequency Filters Fourier transform
- Fourier transform (FT)
- decompose an image into its sine and cosine
components - transform real space images into Fourier or
frequency space images - In a frequency space image, each point represents
a particular frequency contained in the real
domain image.
29Frequency Filters Fourier transform
- Discrete Fourier transform (DFT)
- Inverse DFT
30Frequency Filters Fourier transform
FT (log)
31Frequency Filters
- Basic steps for filtering in the frequency domain
32Frequency Filters
- Frequencies in an image correspond to the rate of
change in pixel values - High frequencies
- rapid changes of gray level values
- Low frequencies
- slow changes of gray level values
33Frequency Filters
- Lowpass filters
- attenuate high frequencies while passing low
frequencies - Highpass filters
- attenuate low frequencies while passing high
frequencies
34Frequency Filters lowpass filters
- Ideal lowpass filters (ILPF)
35Frequency Filters lowpass filters
- Butterworth lowpass filters (BLPF)
36Frequency Filters lowpass filters
- Gaussian lowpass filters (GLPF)
37Frequency Filters highpass filters
- Highpass filters
- Ideal higpass filters (IHPF)
- Butterworth highpass filters (BHPF)
- Gaussian highpass filters (GHPF)
38(No Transcript)
39Frequency Filters bandpass filters
- Bandpass filters
- attenuate very low frequencies and very high
frequencies -
- enhance edges while reducing the noise at the
same time
40Frequency Filters
- Examples (lowpass filters)
ILPF with cut-off frequency of 1/3
ILPF with cut-off frequency of 1/2
BLPF with cut-off frequency of 1/3
BLPF with cut-off frequency of 1/2
Gaussian noise
Original
41Frequency Filters
- Examples (highpass filters)
42Frequency Filters
- Relationship and comparison with spatial filters
- spatial filtering
- frequency filtering
-
43Frequency Filters
- Comparison with spatial filters
- more computational efficient
- more intuitive
44Summary
- Filtering is the operation of applying a
transform on an image in order to enhance it. - Filtering techniques can be subdivided into two
types - Spatial domain filtering
- Frequency domain filtering
45Summary
- Filtering techniques are very useful in image
analysis and processing - Noise removal
- Edge detection
46Thank you Questions ?