Lecture 5: Fundamentals of Image Processing

Question 1

What captures a lot of information?

Answer 1

Images

Question 2

What are the typical sizes?

Answer 2

320x240
640x480
1280x720

Question 3

What size is QVGA?

Answer 3

320 x 240

Question 4

What size is VGA?

Answer 4

640 x 480

Question 5

What size is HD?

Answer 5

1280 x 720

Question 6

What is filtering?

Answer 6

Certain components are accepted or rejected.

Question 7

What does a low-pass filter do?

Answer 7

It smooths and image(allows low frequency)->blurring(smoothing) effect on an image used to reduce image noise.

Question 8

What does a high-pass filter do?

Answer 8

It retains the contours(also called edges) of an image(high frequency)

Question 9

What is the motivation of Low-Pass filtering?

Answer 9

noise reduction

Question 10

What is salt and pepper noise? What is it also known as?

Answer 10

It presents itself as sparsely occurring white and black pixels sometimes seen on digital images.
It is also known as impulse noise.

Question 11

What is gaussian noise?

Answer 11

It has variations in intensity drawn from a Gaussian normal distribution.

Question 12

What are 2 example of Low-pass filtering?

Answer 12

-Salt and pepper noise
-Gaussian noise

Question 13

How do you move average in 1D?

Answer 13

You replace each pixel with an average of all the values in its neighborhood.

Question 14

How do you move average in 2D?

Answer 14

You replace each pixel with an average of all the values in its neighborhood.

Question 15

What is the motivation of High-Pass filtering?

Answer 15

edge detection, an idealized line drawing

Question 16

What do edge contours in the image correspond to ?

Answer 16

important scene contours

Question 17

What are Edges?

Answer 17

Edges are sharp intensity changes.

Question 18

How can edges be represented as functions?

Answer 18

f(x,y)

Question 19

How can edges look like?

Answer 19

steep cliffs

Question 20

An edge is a place of what?

Answer 20

a place of rapid change in the image intensity function

Question 21

What is a fact about images?

Answer 21

They are noisy.

Question 22

What are 4 examples of image noises?

Answer 22

Light Fluctuations
Sensor Noise
Quantization effects: from input values to quantized values
Finite Precision

Question 23

When it comes to the summary on Smoothing Filters, what values does it have?

Answer 23

positive values

Question 24

When it comes to the summary on Smoothing Filters, what does the “low-pass” filter remove?

Answer 24

It removes “high-frequency” components.

Question 25

When it comes to the summary on Derivative Filters, what signs does it have and what are they used for?

Answer 25

It has opposite signs that are used to get high response in regions of high contact.

Question 26

When it comes to the summary on Derivative Filters, what does the “high-pass” filter highlight?

Answer 26

It highlights “high-frequency” components

Question 27

What do you do in non-maximal suppression?

Answer 27

You identify local maxima along a gradient direction.

Question 28

What’s involved in the Canny edge-detection algorithm?

Answer 28

-computing gradient of smoothed image in both directions
-discarding pixels whose gradient magnitude is below a certain threshold
- non-maximal suppresion

Question 29

What is the process of the Canny edge-detection algorithm?

Answer 29

-Take a grayscale image. If not grayscale, convert it into a grayscale by replacing each pixel by the mean value of its R,G,B components.
-Convolve the image with x and y derivatives of Gaussian filter
-Threshold it(set to 0 when the pixel value is below a given threshold)
-Take local maximum along gradient direction.

Question 30

What is Thinning also known as?

Answer 30

Non-maxima suppression, local-maxima detection along edge direction

Question 31

How do you build a panorama? 4 steps

Answer 31

-match/align images
-detect feature points in both images
-find corresponding pairs
-use these pairs to align images

Question 32

What is problem 1 when it comes to point features?

Answer 32

Detecting the same points independently in both images, if they are in the field of view.
There is no chance to match!

Question 33

What do you need to solve problem 1?

Answer 33

a repeatable feature detector

Question 34

What is problem 2 when it comes to point features?

Answer 34

For each point, identify its correspondence in the other images

Question 35

What do you need to solve problem 2?

Answer 35

a reliable and distinctive feature detector

Question 36

What’s true with some patches?

Answer 36

Some patches can be localized or matched with higher accuracy than others

Question 37

Information can be what two things?

Answer 37

Useful or Redundant

Question 38

How do you get useful information from images?

Answer 38

Selecting features that:
are distinctive
do not vary much in appearance
can be detected & matched very fast

Question 39

What’s an example of useful information?

Answer 39

using point features like corners

Question 40

What is the property of corner detection?

Answer 40

shifting a window in any direction should give a large change of intensity in at least 2 directions

Question 41

When is corner detection identified as a “flat” region?

Answer 41

When there is no intensity change

Question 42

When is corner detection identified as an “edge”?

Answer 42

when there is no change along the edge direction

Question 43

When is corner detection identified as a “corner”?

Answer 43

when there is significant change in at least 2 directions.