Lecture 4: Computer Vision Flashcards
What is the most powerful sense?
Vision
The retina has millions of what with a data-rate of what?
photoreceptors with a data-rate of 3 GBytes
Name 5 applications of Computer Vision.
AI, ML, Psychology NeuroSci, Robotics, Autonomous driving, hazardous waste clean-up, search and rescue, space exploration
What does automatic extraction mean?
‘extracting’ meaningful info from images and videos
What are the two types of info that can be extracted and provide examples for both.
Semantic Info(ground, outdoors, shade, roof, sky, door, building)
Geometric Info(shapes)
What are three examples of Automatic Reconstruction and Recognition?
Partial 3D from overlapping images
Dense 3D surface model
Face recognition
What reduces blurring in image formation?
Blocking most of the rays by adding a barrier between the object and the photoreceptive surface.
What is the opening in image formation called?
an Aperture
The depth of the room/box is what?
the effective focal length
What is the pinhole model? What’s the name of the point? Where is that image formed?
captures beams of rays through a single point.
that point is the Center of Projection or Optical Center.
the image is formed on the Image Plane
What are the effects of shrinking the aperture?
The image could suffer from diffraction effects because less light gets through, which increases the exposure.
What is the ideal pinhole?
There’s only one ray of light that reaches each point on the film.
What could make images more blurry?
making the pinhole bigger
Perspective is what?
the dependence of the apparent size of an object on its depth
What is Stereo Vision?
It uses 2 cameras with known relative position T and orientation R to recover the 3D scene information
