Set 1


SET 1

Night vision stereo sequences provided by Daimler AG in June 2007

SEQUENCE LENGTH (FRAMES) 250 / 300
COLOUR / GREY Grey
COLOUR DEPTH 12-bit
RESOLUTION (PIXELS) 640 x 481
STEREO RECTIFIED Yes
EGO-MOTION INFO Yes (in header)
  • Ego-motion info is found in the PGM header: Yaw Rate, Speed and Wheel Rotations for the ego-vehicle. It also contains the frame rate (cycle time). See this report for more details.
  • Intrinsic and extrinsic camera parameters are available for download.

Sample Images & Downloads

  • Videos of the sequence along with trajectories of the ego-vehicle (by Ali Al-Sarraf) can be seen by clicking on the respective image.
  • Each sequence is around 70MB (250 Frames) or 80MB (300 Frames), for each camera sequence.
SEQUENCE NAME TOTAL FRAMES
Construction-Site 300
LEFT CAMERA RIGHT CAMERA
ConstructionSite-left.7z ConstructionSite-right.7z
SEQUENCE NAME TOTAL FRAMES
Crazy-Turn 250
Crazy turn sequence
LEFT CAMERA RIGHT CAMERA
CrazyTurn-left.7z CrazyTurn-right.7z
SEQUENCE NAME TOTAL FRAMES
Dancing-Light 250
Crazy turn sequence
LEFT CAMERA RIGHT CAMERA
DancingLight-left.7z DancingLight-right.7z
SEQUENCE NAME TOTAL FRAMES
Intern-On-Bike 250
Crazy turn sequence
LEFT CAMERA RIGHT CAMERA
InternOnBike-left.7z InternOnBike-right.7z
SEQUENCE NAME TOTAL FRAMES
Safe-Turn 300
Crazy turn sequence
LEFT CAMERA RIGHT CAMERA
SafeTurn-left.7z SafeTurn-right.7z
SEQUENCE NAME TOTAL FRAMES
Squirrel 300
Crazy turn sequence
LEFT CAMERA RIGHT CAMERA
Squirrel-left.7z Squirrel-right.7z
SEQUENCE NAME TOTAL FRAMES
Traffic-Light 250
Crazy turn sequence
LEFT CAMERA RIGHT CAMERA
TrafficLight-left.7z TrafficLight-right.7z

You can also download the stereo camera intrinsic parameters text file here: camera.dat


Some work done with these sequences:

  1. Introductory report, explaining the given data (Zhifeng Liu and Reinhard Klette, October 2007)
  2. Early (2007) experimental results using those sequences
  3. Belief-propagation
  4. Talk about evaluation results at Dagstuhl/Germany (Reinhard Klette, July 2008)
  5. Approximate ground truth, dynamic programming, etc. (Zhifeng Liu and Reinhard Klette, 2008)
  6. 3D scene flow estimation (Jorge A. Sanchez, Reinhard Klette, and Eduardo Destefanis, 2008)
  7. Semi-global matching (Simon Hermann, Reinhard Klette and Eduardo Destefanis, 2008)
  8. Edge detectors, belief propagation, approximate ground truth, etc. (Shushi Guan, Reinhard Klette and Young Woon Woo, 2008)

Reference:

Tobi Vaudrey, Clemens Rabe, Reinhard Klette, and James Milburn, “Differences Between Stereo and Motion Behaviour on Synthetic and Real-World Stereo Sequences“, in Proc. 23rd Int. Conf. Image and Vision Computing New Zealand (IVCNZ ’08), 1-6, 2008. (Click here to download this as a bib file).