Context-Sensitive Map GPS and Video

1
Context-Sensitive Map GPS and Video

• Motivation
• Highway patrol and paramedics could benefit from
  increased range of vision when managing traffic
  and viewing highway accident scenes.
• Drivers tend to navigate using familiar landmarks
  (such as buildings), in addition to using street
  names and numbers.

Edward Shyu Computer Vision and Robotics Research
(CVRR) Lab Advisor Dr. Mohan Trivedi Sponsored
by Cal (IT)2 Intelligent Transport and
Telematics

• Description
• Pan-tilt-zoom cameras placed around campus and
  along highways send video data through the
  internet.
• Wirelessly enabled laptop retrieves video
  information.
• GPS receiver connected to laptop allows programs
  automatic camera selection to choose appropriate
  camera to display and adjust.

• Goal
• Program becomes an extension of users eyes,
  allowing user to see a much wider range of the
  surrounding area.
• Conventional map-GPS users may navigate by
  looking for landmarks (buildings, parks) that can
  be previewed using the program, thereby making
  navigation more intuitive

pan-tilt zoom cameras
Camera positions are denoted by colored dots.
Selected camera is hi-lighted in yellow
Real-time video from pan-tilt-zoom cameras
stationed on campus
User can adjust orientation and zooming of
selected camera
laptop computer, wireless internet
Choose specific camera to view and adjust, or
allow program to select appropriate
camera. Currently, AutoCam chooses closest
camera algorithm may also choose cameras that
are along the path in which the user is traveling.
GPS receiver
MapCam program
Aerial still images provide more detail to aid
navigation.
User location is denoted by smiley icon on map.
Coordinates are displayed below map.