Computer Vision Flashcards

Perspective projection for pinhole camera Graph+All names 9

1

Find equations for scene points x,y → image points u,v 9

1

Find disparity equation for stereo system (graph + demo) 11

1

Explain what an epipole is and why we rectify them 12

1

Define vanishing point and explain how to find it 13

1

Explain why we use lenses 15

1

Give thin lenses equations and equation fur u and v 15

1

Explain what circles of confusions are 16

1

Define the depth of field 17

1

explain why we use a diaphragm 17

1

Define F numbers 17

1

Explain how autofocus works 17

1

Define sampling and quantization+ eq for memory size of picture 18

1

Give the two main sensors technologies for cameras 19

1

Give PRO/CON for CMOS and CCD technologies 19

1

Explain two ways to denoise images 20

1

Define SNR and give equation (relation with parameters) 20

1

Give four noise sources and explain them 20

1

Define dynamic range + equation 21

1

Define High dynamic range 21

1

Explain how to achieve color images + impact on resolution 22

1

Explain RGB /HSI/LAB color spaces 22

1

Explain what projective space is 23

1

How are points at infinity represented in projective space ? Use line example 23

1

Give perspective projection linking m_tilde to M_tilde threw P_tilde 25

1

Define perspective projection matrix P_tilde 26

1

Define cannonical/standard PPM 26

1

Give demo of vanishing points using projective space 26

1

Giive the two issues we have to solve for a good camera model 26

1

Explain transformation linked to image digitalization (equations) 26

1

Define intrinsic parameter matrix + link with PPM 27

1

define WRF and CRF 27

1

demo to get relation between m_tilde and W_tilde 27

1

Give general PPM form 27

1

Define extrinsic parameter matrix 28

1

Define homographies and give 4 possible configurations with graph and proof 28

1

Explain two main effects of lens distortions 28

1

Give equation of lens distortion + approximation tyler 28

1

Give the image formation flow 29

1

Define calibration and the two main approaches 30

1

Draw the zang method flowchart + basic assumptions 30

1

Give P as a homograpphy + nbr éq needed 31

1

Give DLT algorithm used to determine H via SVD 32

1

Explain SVD method 33

1

Explain the general way to obtain algebraic vs geometric error 33

1

Give equation for H using Least Square reprojection error 33

1

Do demo to find intrinsic parameters based on H 35

1

Do the demo to find extrisic parameters 36

1

Do demo to find lens distortion coeficients 37

1

Explain how to raffine using repropjection error (A, R and T ) with the name of the algorithm 37

1

Define what a wrapping is 37

1

Define forward / backward mapping + pros/cons 38

1

Explain + equation billinear mapping 39

1

Explain the naive way of calibrating a stereo camera 37

1

Explain demo to find hopmographies with stereo camera calibration (New PPMs and homographies) 37

1

Démo + éq to retreive 3D coord of stereo camera data 37

1

Explain why gray level histograms are usefull 40

1

Define LTE in spacial filtering 46

1

Equation for LTE linearity and translation equivalence 46

1

Define the impulse response 47

1

Give the equation of the 2D convolution + 4 propreties 48

1

Visual representation of a 2D concolution 48

1

Equation of correlation + 4 propreties 49

1

Visual representation of 2D correlation 49

1

3 propreties of correlation and convolution relations 50

1

Give the discrete equation of convolution + how to compute it 50

1

What happens ate the borders when using a convolution ? 51

1

Explain mean filtering + convolution method + kernel size tradeoff 51

1

Define size and scale 51

1

Explain gaussian filtering + convolution method + variance vs size usual rule of thumb 53

1

Explain median filter and give an example of application 55

1

In wich order to perform median / linear filtering ? why ? 56

1

Explain and give equations of the billinear filter 56

1

Explain and give the equations of the non local mean filte 57

1

Expplain image binarization 59

1

Explain intensity thresholding + limitation 60

1

Give 3 automatic threshold selection method 61

1

Explain equations of otsu’s threshold algorithm 62

1

Exoplain adaptive thresholding and why it is needed 64

1

Explain coulour bases segmentation + equations 64

1

Why is eucidian distance not enough for color segmentation ? 65

1

define Mahalanobis distance + equations + justification of diagonalisation 67

1

Explain dilation, erosion, opening and closing 69

1

Define BLOBS 74

1

Define 4-neighbours / 8-neighbours connectivity 75

1

Define connected regions 76

1

Explain blob area / barycentre / Perimeter / orientation 81

1

Explain blob elongatedness / Rectangularity / ellipticity 89

1

Define an edge 92

1

Explain simple threshold on derivative method (graph+eq) 92

1

Define with graph and equation backward / central / forward discrete gradient approximation 94

1

Why is noise a problem for edge detectors ? 95

1

Explain with equations and graphs the smoothing derivative edge dectors operators 95

1

Explain prewitt and sobel edge detector operators + équations 96

1

NMS : concept + graph with gradient + pixel visualization + flowchart 97

1

Define the laplacian + discrete operations + kernel 99

1

Explain LOG edge detector / 3 steps / sigma impact / final thresholding 100

1

3 criteriums for a good edge detector 102

1

Canny’s edge detector / concept / flowchart / result 104

1

Give the three steps opf establishing correspondances in images 106

1