Lecture 4 Flashcards

1
Q

Image Filtering

A

techniques for modifying or enhancing an image

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

Types of Image Filtering

A

Linear filters and Non-linear filters

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

Which are the linear filters

A

image smoothing/blurring
edge detection
image sharpening

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

Which are the non-linear filters

A

morphological operators (dilation,erosion)
median filter (noise reduction)
matched filters

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

the value in each output pixel depends on the which values

A

the local values of neighboring pixels

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

what does a non-linear filtering

A

replace each pixel by a non-linear combination of its neighbors

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

what does linear filtering

A

replace each pixel by a linear combination of its neighbors

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

Image Filtering Steps

A
  1. move filter matrix over the image
  2. multiply the corresponding image pixel
  3. sum the results and store it as a new pixel value
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

Image smoothing

A

replace each pixel with the average intensity of its neighbors

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

cross-correlation

A

measures the similarities between two signals

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

convolution

A

measures the similarities between two signals, but the filter is flipped upside-down

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

which convolutions can save a lot of time

A

separable convolutions

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

strides

A

skip steps at a time (for efficiancy)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

padding

A

extending the frame of an image with empty pixels, so when the kernel goes over the image we can maintain the size of the image

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

why we need padding

A

allow for more space for the kernel to cover the whole image

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

blurring

A

better image smoothing

17
Q

Filters used for better image filters

A

gaussian filters

18
Q

what are doing the gaussian filters

A

remove the high-spatial frequencies

19
Q

edge detection

A

edges are sudden transition of intensity and are very informative
laplacian: measures the 2nd derivative
sensitive to noise
2 filters => sober and robert filters
subtract two blurred versions of the same image

20
Q

Laplacian of Gaussian-filtered image (LoG)

A

measure of the 2nd derivative (the change) of the image

21
Q

What is shown by the 2nd derivative in Laplacian?

A

showing the magnitude of a change

22
Q

What does the output of Laplacian stands for?

A

large and positive on one side of the edge
large and negative on the other
small/zero at the edge itself

23
Q

image sharpening

A

emphasize details of images:
edges and local contrast

24
Q

what is image sharpening

A

emphasize the details of an image

Image + alpha*EdgeDetection

25
Q

dilation

A

takes the max value of all pixels in its neighborhood

26
Q

erosion

A

takes the min value of all pixels in its neighborhood

27
Q

non-linear morphological operators

A

dilation
erosion

28
Q

what does the median filter

A

noise reduction

29
Q

median filter

A

replace each pixel with the median value within its neighbor

30
Q

common models for noise reduction

A

gaussian additive noise
salt-pepper noise

31
Q

Mexican Hat Filter

A

indicates the shape obtained by subtracting two Gaussians (“Difference of Gaussians”) with different standard deviations.

Because “Mexican hat” refers to the shape it is sometimes also used when referring to the shape of the Laplacian filter which is very similar to that of the DoG filter.