Download presentation
Presentation is loading. Please wait.
Published byLee Davidson Modified over 6 years ago
1
Basis beeldverwerking (8D040) dr. Andrea Fuster dr. Anna Vilanova Prof
Basis beeldverwerking (8D040) dr. Andrea Fuster dr. Anna Vilanova Prof.dr.ir. Marcel Breeuwer Convolution
2
Contents Spatial filtering Correlation Convolution Filters:
Smoothing filters Sharpening filters Borders Basis beeldverwerking 8D040
3
Spatial filtering Input image , use a filter to obtain processed image
Filter consists of Neighbourhood (rectangular) Mostly odd dimensions Predefined operation Create new pixel value in center of neighbourhood Basis beeldverwerking 8D040
4
Spatial filtering Filter operation (3x3 filter)
More compact notation filter Basis beeldverwerking 8D040
5
Intuition to filtering
Basis beeldverwerking 8D040
6
Move filter over image Basis beeldverwerking 8D040
7
Basis beeldverwerking 8D040
8
Basis beeldverwerking 8D040
9
Basis beeldverwerking 8D040
10
Basis beeldverwerking 8D040
11
Correlation While moving the filter, at each position
Multiply values of overlapping locations Sum all multiplication results Basis beeldverwerking 8D040
12
Correlation vs. Convolution
Discrete Correlation 2D Discrete Convolution 2D - Equivalent to first rotate the filter 180 degrees and correlate- Basis beeldverwerking 8D040
13
Example See blackboard ☺ (or figure 3.30 Gonzalez and Woods)
Basis beeldverwerking 8D040
14
Correlation vs. Convolution
Correlation and convolution give the same result if the filter used is symmetric! Basis beeldverwerking 8D040
15
Convolution – 1D cont. case
Imagine a system with input signal transfer function output signal then Basis beeldverwerking 8D040
16
system transfer function
Definition input output system transfer function Basis beeldverwerking 8D040
17
Dirac delta function (unit impulse)
Definition Constraint Sifting property Specifically for t=0 Basis beeldverwerking 8D040
18
Convolution Let We saw this already in the discrete case
Basis beeldverwerking 8D040
19
Properties of convolution
Commutative Associative Distributive Basis beeldverwerking 8D040
20
Convolution is commutative
Proof Let Q.E.D. Basis beeldverwerking 8D040
21
Convolution is associative - 1
Proof Basis beeldverwerking 8D040
22
Convolution is associative - 2
Basis beeldverwerking 8D040
23
Convolution is associative - 3
Let Basis beeldverwerking 8D040
24
Convolution is associative - 4
Q.E.D. Basis beeldverwerking 8D040
25
Convolution is distributive - 1
Proof Basis beeldverwerking 8D040
26
Convolution is distributive - 2
Q.E.D. Basis beeldverwerking 8D040
27
Discrete convolution 1D 2D Basis beeldverwerking 8D040
28
Discrete convolution Formulas take summation from to
Filters have a limited size, e.g., 1D a + 1 2D (2a + 1, 2b + 1) Basis beeldverwerking 8D040
29
Filter Kernels Kernel Basis beeldverwerking 8D040
30
Filters Idea: correlate or convolve image with different filters in order to obtain different results, i.e., processed images Basis beeldverwerking 8D040
31
Smoothing filters Average intensities – result is blurred image, less details Response: (z’s image intensities) … NxN filter Basis beeldverwerking 8D040
32
Smoothing filters Note that:
Sum of filter coefficients is 1 (normalized filter) Correlation = convolution (symmetric filter) Filter size effect? Basis beeldverwerking 8D040
33
Smoothing filters - example
Original x3 smoothing filter NxN filter (see figure 3.33 in Gonzalez and Woods!) Basis beeldverwerking 8D040
34
Effect of normalized smoothing kernel
non- normalized Basis beeldverwerking 8D040
35
Sharpening filters Enhance parts of the image where intensities change
rapidly, such as edges Basic derivative filters Measure change of intensity in x or y direction Basis beeldverwerking 8D040
36
Example Basis beeldverwerking 8D040
37
Arbitrary angle derivative
Given and Basis beeldverwerking 8D040
38
Arbitrary angle derivative
Basis beeldverwerking 8D040
39
Prewitt gradient kernel
Derivative in one direction, smoothing in the perpendicular direction Basis beeldverwerking 8D040
40
Example Prewitt Basic derivative Basis beeldverwerking 8D040
41
Sobel kernel Basis beeldverwerking 8D040
42
Example (Thanks to Wikipedia☺)
43
Derivative filters Note that coefficients in all of the previous filters sum to zero, i.e., zero response in area of constant intensity Also: gradient, Laplacian, … Basis beeldverwerking 8D040
44
Borders Do you see any problems at image borders? Try position (0,0)
Basis beeldverwerking 8D040
45
Border problems How to handle? No border handling Padding
Border is not filtered Padding Put values outside image border Cyclic padding Use values from the other side of the image Basis beeldverwerking 8D040
46
Zero padding Basis beeldverwerking 8D040
47
Cyclic padding Basis beeldverwerking 8D040
48
Padding Remember: padding is artificial!
The values chosen outside the border influence the outcome image Basis beeldverwerking 8D040
49
End of part 2 Thanks and see you Tuesday 21!
Basis beeldverwerking 8D040
Similar presentations
© 2024 SlidePlayer.com. Inc.
All rights reserved.