Computer Science 455
Instructor: R. P. Burton
Third Quiz
October 10-11, 2005
Name _________________________________________ Score ____________/26
Suppose a complicated image has been scan-converted into a rectangular region of the frame buffer, but needs to be translated into another, possibly overlapping region of the frame buffer. Which approach is best:
Re-scan-convert the image in the new location
move one scan-line at a time, starting with a corner of the scanned image that is not overlapped.
Move one scan-line at a line, starting with a corner of the destination image that is not overlapped.
Move one scan-line at a line, starting with an overlapped corner.
(d)
Your data consists of eight variables. You wish to scale the data. How many entries in the scaling matrix consist on nonzero values?
8
9
16
17
18
more than 18
(d) (assuming the scaling takes place about an arbitrary point; (b) otherwise)
Suppose you wish to rotate (in 3D) about a line segment which determines an arbitrary axis. To calculate the sines and the cosines for the first angle of rotation, you should use the length of
the line segment
a projection of the line segment
a projection of the projection of the line segment
(b)
y points out, x points left, z points down. To rotate counterclockwise about y, the minus sign on the sine goes in the _____ row (where the top row is the “first” row).
first
second
third
fourth
(a)
For three-dimensional data, reflections can occur relative to any of the following EXCEPT
a point
a line
a plane
(no exceptions here)
(d)
Suppose the data is represented relative to 3D coordinate system A, but we would like to represent it relative to 3D coordinate system B. Which of the following is the correct approach?
find a transformation which maps from A to B and apply it to the data
find a transformation which maps from A to B and apply it to the coordinate frame A
find a transformation which maps from B to A and apply it to the data
find a transformation which maps from B to A an apply it to coordinate frame A
(c)
3D data can be sheared with respect to (pick the most inclusive answer)
one coordinate
any two coordinates, though no coordinate can be “sheared” relative to itself
any two coordinates
any three coordinates
(d)
In traditional (i.e. early) graphics, a window is an area in
display device space
world coordinate space
(b)
The relationship of a viewport to its space is identical to the relationship of a window to its space.
true
false
(b)
Which makes the most sense?
rotated window
a rotated viewport
both make (roughly) equivalently good sense
neither makes much sense
(a)
A change to a viewport produces _______ change in the scene; a change to a window produces _____ in the scene.
a comparable, a comparable
a comparable, a inverse
an inverse, a comparable
an inverse, an inverse
(b)
The window-to-viewport mapping (with no rotation)
is a direct mapping requiring no transformations
requires only a translation
requires only a scaling
requires a scaling and a translation
cannot be represented in matrix form (at least based on what has been discussed in class)
(d)
In an affine transformation
parallel lines remain parallel
angles are preserved
lengths are preserved
(all of the above)
(none of the above)
(a)
All things considered, it is more efficient to clip to the
viewport
window
(it makes no difference)
(a)
If a window is rotated, the conventional clipping algorithms do not apply.
true
false
(b)
___________ is/are intended only for clipping regions with 2 vertical boundaries and 2 horizontal boundaries.
Sutherland-Hodgman
Weiler-Atherton
(both (a) and (b))
(a) (Sutherland-Hodgman can be adapted to work for any convex clipping polygon, but as presented in the text and other sources, it is for grid-aligned rectangles only)
The Cohen-Sutherland algorithm is like playing tic-tac-toe; you try to get 3 “X”s in a line.
true
false
(b)
The first test/activity in Cohen-Sutherland is
attempt a trivial acceptance
attempt a trivial rejection
divide along a boundary
(a)
The number of subdivisions in Cohen-Sutherland with midpoint subdivision is proportional to
the length of the line segment
the square of the length of the line segment
the square root of the length of the line segment
the log of the length of the line segment
(d)
Liang and Barsky can determine all of the following except
a line segment parallel to a boundary
a line segment external (or internal) to a boundary
whether a line segment enters or exits a boundary
where a line segment passes through a boundary
(no exceptions here)
(e)
The primary activity of Nicholl Lee Nicholl is
determining if a line segment is parallel to a boundary
determining if a line segment passes through a boundary
determining where a line segment passes through a boundary
(b)
A clipping algorithm should be designed to accept polygons and produce _____.
points
lines
points and lines
polygons
polygons and lines
polygons, lines and points
(d)
_________ can produce extraneous lines (which are not part of the original scene).
Cohen-Sutherland
Sutherland-Hodgman
Weiler-Atherton
(all of the above)
(b)
To clip text (aligned with the vertical or horizontal boundaries), “all or nothing”
clip the major diagonal of the text’s bounding box
clip the minor diagonal of the text’s bounding box
((a) XOR (b), depending on the position of the text’s bounding box
(a) or (b); both always work
(d)
Curve clipping
can be done using the algorithms referenced above
is beyond the scope of CS 455
(a)
Blanking
can be done using the algorithms referenced above
is beyond the scope of CS 455
(a)