CV Chapter 2 Image Processing Basics

Questions about the lecture 'Computer Vision' of the RWTH Aachen Chapter 2 Image Processing Basics

106

0.0 (0)

R. B.

Étudier

R. B.

À propos de ce fichier

This flashcard set covers fundamental concepts of image processing at a university level. It delves into edge detection, filters, and multi-scale representations, focusing on techniques like the Hough transform, chamfer distance, and various edge detection methods such as the Canny edge detector. The set also explores image gradients, Laplacian operators, and finite difference filters, providing practical MATLAB code examples. Ideal for computer science students and professionals, this flashcard set offers a comprehensive overview of essential image processing tools and their applications.

Cartes-fiches

106

Utilisateurs

Langue

Anglais

Catégorie

Informatique

Niveau

Université

Créé / Mis à jour

04.02.2017 / 18.02.2017

Fichiers de cartes similaires

What is the MATLAB code for gaussian smoothing? [6]

>> for sigma=1:3:10

>> h = fspecial('gaussian‘, fsize, sigma)

>> out = imfilter(im, h);

>> imshow(out);

>> pause;

>> end

What is the MATLAB code for canny edge detector? [2]

>> edge(image,‘canny’);

>> help edge

What types of fitting as structure extraction exist? [2]

1. Template matching and 2. parametric search

What are the characteristics of template matching in fitting? [2]

1. Associate model (line, circle, shape) with observed features // Similar as template matching with edge templates (Traffic sign detection)

2. Correlation does not work because of value zero if one pixel off -> Use chamfer distance, distance transform

What are the characteristics of chamfer distance? [4]

1. Average distance to nearest edge pixel

2. D_Chamfer(x,y) = 1/T * Sum_[T(u,v)=1] dt(x+u,y+v)

3. More robust

4. Replace Kernel T by fixed number of points T

What are the advantages of chamfer distance? [3]

1. Fast and simple

2. Matches upright shapes with little intraclass variation

3. Matches candidate in longer recognition pipeline

What are the disadvantages of chamfer distance? [3]

1. Not discriminative and needs further verification

2. Many false positives detections for cluttered regions

3. Expensive for rotated rescaled objects

What is the definition of distance transform in 1D? [3]

1. Initialize with 0 and inf

2. Forward taking min(self, left+1) // Missing first

3. Backward taking min(self, right+1) // Missing last

What has to be changed for distance transform in 2D?

Replace left and right pixel above or left and below or right

What are the characteristics of parametric search in fitting? [4]

1. Parametric model with observed features

2. Infeasible to check all combinations of models

3. Membership criteria is not local

4. Hough transform (HT) with hough space (HS)

What is the idea of hough transform?

Transform edge pixels into lines and vote for them to deduce the best lines

How are elements of the image space saved in hough space? [3]

1. Save point (m,b) in HS for each line y=m*x+b // Undefined for vertical lines

2. Save line b=x*m+y in HS for each point (x,y)

3. Save (d,t) in accumulator array H with x*cos(t)-y*sin(t)=d // Polar representation

How can the elements in hough space be used? [2]

1. Line by p1 p2 in IS is crossing of two lines in HS

2. Vote for best line by more than two points

What is the hough transform algorithm? [4]

1. Init H(d,t) with 0

2. Forall edge pixel (x,y) and t=0 → 180 increase H(d,t)

3. Find maximizing (d,t)

4. Return line x*cos(t)+y*sin(t)=d

What are the extensions of the the hough transform algorithm? [4]

1. t=G in 2.

2. Give more vote for stronger edges

3. Change resolution of H

4. Change shape

What the characteristics of a circle used as shape in hough transform algorithm? [3]

1. (x-a)²+(y-b)²=r with fix r

2. For unkown r and G with cone (Unkown G) or line (Known G) in 3D

3. forall r t (Estimated with G or each): a=x-r*cos(t) and b=y+r*sin(t)

What are the notes for hough transform? [4]

1. Remove irrelevant pixels

2. Vote for neighbors as smoothing in H

3. Use direction of edge to reduce free parameters

4. Tag votes

What are the advantages of hough transform? [3]

1. Independent pixel process

2. Robustness to noise

3. Detection of multiple instances in a single pass

What are the disadvantages of hough transform? [3]

1. Fast increasing complexity

2. Not relevant peaks in parameter space

3. Picking good grid size

What are the characteristics of generalized hough transform? [2]

1. Detect arbitrary shape by boundary points P:={p1,…, pn} in IS

2. Compute displacement vectors r=a-pi and store in table indexed by G orientation t

Which values have pixels of a binary image?

Commandes clavier:

= tourner,

= avant/arrière,

= faire défiler

Zero for white and one for black // regions of interest

What are binary images used for? [9]

Commandes clavier:

= tourner,

= avant/arrière,

= faire défiler

1. industrial inspection

2. document analysis

3. text recognition

4. medical/bio data

5. blob tracking

6. motion analysis

7. shape analysis

8. free viewpoint video

9. detection

What is the idea of detection with binary images?

Commandes clavier:

= tourner,

= avant/arrière,

= faire défiler

pixel → 0/1 dependent on intensity value or color range

What are the issues of binary images? [2]

Commandes clavier:

= tourner,

= avant/arrière,

= faire défiler

1. multiple regions of interest

2. noisy binary output

What is the definition of conversion to binary images?

Commandes clavier:

= tourner,

= avant/arrière,

= faire défiler

Map every pixel to black or white dependent on a threshold

What variants of conversion to binary images exist? [3]

Commandes clavier:

= tourner,

= avant/arrière,

= faire défiler

1. One sided, 2. two sided and 3. membership

What variants of threshold selection exist? [2+1]

Commandes clavier:

= tourner,

= avant/arrière,

= faire défiler

1. Maximize sigma_between by Otsu (1979)

2. Local binarization by Niblack (1986)

3. Detect smooth gradient and approximate with polynomial surface fitting

What noise has to be removed?

Commandes clavier:

= tourner,

= avant/arrière,

= faire défiler

Isolated points and small structures

What noise has to be filled?

Commandes clavier:

= tourner,

= avant/arrière,

= faire défiler

Holes

Name five morphological operators? [5]

Commandes clavier:

= tourner,

= avant/arrière,

= faire défiler

1. Dilation, 2. erosion, 3. opening, 4. closing and 5. boundary extraction

What is the idea of morphological operators?

Use structuring element (SE) to analyze image content

How are morphological operators defined on gray scale images?

Use neighboring max or min

What are the characteristics of dilation? [2]

1. If pixel is foreground set all pixels to foreground

2. Expand connected components, grow features and fill holes

What are the characteristics of erosion? [2]

1. If pixel is not foreground set all pixels to background

2. Erode connected components and shrink features

What are the characteristics of opening? [2]

1. Keep shape and remove small objects

2. Feature selection through shape

What are the characteristics of closing?

Keep shape and fill holes

What are the characteristics of boundary extraction? [1]

Get 1 pixel thick boundary line with 3x3 SE

What is the definition of dilation?

Change black SE center pixel into white if one of the SE pixels covers a white pixel

What is the definition of erosion?

Change white center pixel into black if one of the SE pixels covers a black pixel

What is the definition of opening?

Perform first erosion then dilation

1 / 106

Cartes-fiches

Which values have pixels of a binary image?

Zero for white and one for black // regions of interest

What are binary images used for? [9]

1. industrial inspection

2. document analysis

3. text recognition

4. medical/bio data

5. blob tracking

6. motion analysis

7. shape analysis

8. free viewpoint video

9. detection

What is the idea of detection with binary images?

pixel → 0/1 dependent on intensity value or color range

What are the issues of binary images? [2]

1. multiple regions of interest

2. noisy binary output

What is the definition of conversion to binary images?

Map every pixel to black or white dependent on a threshold

What variants of conversion to binary images exist? [3]

1. One sided, 2. two sided and 3. membership

What variants of threshold selection exist? [2+1]

1. Maximize sigma_between by Otsu (1979)

2. Local binarization by Niblack (1986)

3. Detect smooth gradient and approximate with polynomial surface fitting

What noise has to be removed?

Isolated points and small structures

What noise has to be filled?

Holes

Name five morphological operators? [5]

1. Dilation, 2. erosion, 3. opening, 4. closing and 5. boundary extraction

What is the idea of morphological operators?

Use structuring element (SE) to analyze image content

How are morphological operators defined on gray scale images?

Use neighboring max or min

What are the characteristics of dilation? [2]

1. If pixel is foreground set all pixels to foreground

2. Expand connected components, grow features and fill holes

What are the characteristics of erosion? [2]

1. If pixel is not foreground set all pixels to background

2. Erode connected components and shrink features

What are the characteristics of opening? [2]

1. Keep shape and remove small objects

2. Feature selection through shape

What are the characteristics of closing?

Keep shape and fill holes

What are the characteristics of boundary extraction? [1]

Get 1 pixel thick boundary line with 3x3 SE

What is the definition of dilation?

Change black SE center pixel into white if one of the SE pixels covers a white pixel

What is the definition of erosion?

Change white center pixel into black if one of the SE pixels covers a black pixel

What is the definition of opening?

Perform first erosion then dilation

What is the definition of closing?

Perform first dilation then erosion

What are advantages of binary images? [3]

1. Fast computation and small storage

2. Simple processing techniques

3. Useful for constrained scenarios

What are disadvantages of binary images? [2]

1. Noise is common in realistic scenarios

2. Can not deal with 3D changes

What are the two types of connectedness?

4 and 8 connected

How does the sequential algorithm for connected components works? [4]

1. If top and/or left have same color copy label

2. If top left have different labels? → copy left label and store connected labels

3. If top left are not labeled → create new label

4. Finally re-label components with smallest label

Which region properties exist? [6]

1. Area A

2. Centroid

3. Circularity C

4. Invariant descriptors with spatial moments

5. Extremal points, bounding box

6. Axis of least second moment

What is the definition of area A?

Count n1 in R

What is the definition of centroid?

Pixel (x’y’) with x’ y’ = 1/A * sum x y for all (xy) in R

What is the definition of circularity C?

1. C = mye_R / sigma_R

2. Mean mye_R= 1/K * Sum_k=0^K-1 |(x_k,y_k) - Centroid|

3. Variance sigma_R² = 1/K * Sum_k=0^K-1 [|(x_k,y_k) - Centroid| - mye_R]²

What is the definition of central moment CM?

mye_jk = Sum_(x,y) (x-x’)^j * (y-y’)^k

Which CMs exist? [4]

0^th CM is area

2^nd CM is variance

3^rd CM is skweness // Schiefe

4^th CM is kurtosis // Wölbung

What is the definition of normalized CM?

peta_pq = mye_pq / mye₀₀^gamma with gamma = p+q/2 + 1

How many invariant moments exist?

Phi1 … phi7

What is the definition of phi1?

phi1 = peta20 + peta02

What is the definition of phi2?

phi2 = (peta20 - peta02)^2 + 4*peta11^2

What is the definition of phi3?

phi3 = (peta30 – 3*peta12)^2 + (3*peta21 - peta03)^2

What is the definition of phi4?

phi4 = (peta30 + peta12)^2 + (peta21 + peta03)^2

What are the characteristics of invariant moments? [3]

1. Robust to translation, rotation and scaling

2. Use log10(phi_i) instead of phi_i directly

3. Do not expected wonders, still summary statistics

What are the motivations for filters? [3]

1. Image restoration

2. Structure extraction

3. Noise reduction

What are the types of noise? [3]

1. Salt and pepper

2. Impulse

3. Gaussian

Étudier

CV Chapter 2 Image Processing Basics

Créer ou copier des fichiers d'apprentissage

Cartes-fiches

Veuillez vous connecter

Google

Apple ID

Microsoft

Switch edu-ID

Edulog

Google

Apple ID

Microsoft

Switch edu-ID

Edulog