CPSC 425 Edges and Corners 20 (Term 2) Practice Questions
Multiple Part True/False Questions. For each question, indicate which of the statements, (A)–(D), are true and which are false? Note: Questions may have zero, one or multiple statements that are true.
Question 1: We have considered two approaches to edge detection, one based on extrema of a 1st derivative operator and the other based on zero crossings of a 2nd derivative operator. Which of the statements, (A)–(D), are true when designing a digital filter to use for differentiation and which are false?
(A) Filter values are normalized to sum to one.
(B) Filter values are normalized to sum to zero.
(C) Filtervaluesarenormalizedsothatthesumofthesquaredvaluesisone(i.e.,sothatthefilter has magnitude one).
(D) Filter values are whatever they are. Normalization is not required.
Solution : (A) False, (B) True, (C) False, (D) False
Question 2: Two thresholds are used when linking edge points in Canny edge detection. Which of the statements, (A)–(D), are true of Canny edge detection and which are false?
(A) Different thresholds are needed to select edge points when linking edges forward or back- ward from the starting location.
(B) The detection of edge points is more accurate when two thresholds are used.
(C) The use of two thresholds prevents gaps that would otherwise appear in the linked edge points.
(D) The X and Y directional derivatives each require a threshold when linking to new edge points.
Solution : (A) False, (B) False, (C) True, (D) False,
Question 3: The Harris corner detector is stable under some image transformations. For which of the image transformations, (A)–(D), is it true that the Harris corner detector is stable? For which is it false? Hint: Features are considered stable if the same locations on an object are typically selected in the transformed image.
(A) Image scaling. (B) Image translation.
(C) Image rotation.
Solution : (A) False, (B) True, (C) True.
Short Answer Questions.
Question 4: Name four scene properties that would cause an edge (brightness discontinuity) in an
• a depth discontinuity (i.e., a foreground/background segmentation)
• a surface orientation discontinuity (e.g., two intersecting planar surfaces)
• a reflectance discontinuity (i.e., a change in surface colour/material on an otherwise smooth surface)
• illumination boundaries (e.g., cast shadows, light sources, specularities) Question 5: Consider the matrix, M, defined at each image point where
Ix2 IxIy M= Ix 2
Note that M can also be written as the outer product of the image gradient, [Ix,Iy], with itself. That is,
M = I [Ix,Iy]
(a) Assuming Ix and Iy are not both zero, what is the rank of M?
Solution: M has rank 1.
(b) Write expressions for the eigenvalues, λ1 and λ2, of M.
λ1 =Ix2+Iy2 λ2 = 0
(c) Is the computation of M at each image point a linear operation? Is it shift invariant? Solution: It is non-linear but it is shift invariant