IMO stereo sequences provided by the European Drivsco project in September 2008.
Sequence Length (Frames) | 850 ~ 1600 |
---|---|
Colour / Grey | Colour |
Colour Depth | 8-bit |
Resolution (pixels) | 1280 x 1024 |
Stereo Rectified | Yes |
Ego-motion info | No |
- Intrinsic and extrinsic camera parameters are available for download.
- IMO (Independent Moving Objects) ground is available for the Suburban Bridge and Follow sequences.
- Gaze data (eye movements of the driver) is available for Suburban Bridge.
- Videos of the sequences (at increased speed) can be seen by clicking on the respective image.
Sample Images & Downloads
Sequence Name | Total frames |
---|---|
Suburban Bridge | 851 |
![]() |
|
Left Camera | Right Camera |
SuburbanBridge-left.7z (1.3 GB) | SuburbanBridge-right.7z (1.3 GB) |
Sequence Name | Total frames |
---|---|
Suburban Follow | 1182 |
![]() |
|
Left Camera | Right Camera |
SuburbanFollow-left.7z (2 GB) | SuburbanFollow-right.7z (2 GB) |
Sequence Name | Total frames |
---|---|
Trailer Follow | 1600 |
![]() |
|
Left Camera | Right Camera |
TrailerFollow1-left.7z (1.3 GB) | TrailerFollow1-right.7z (1.3 GB) |
TrailerFollow2-left.7z (1.3 GB) | TrailerFollow2-right.7z (1.3 GB) |
The camera parameters for the Suburban bridge and Suburban sequences follow can be downloaded from here: Suburban.txt
The camera parameters for Trailer follow are available in here: Trailer.txt
Both parameter files follow the the OpenCV format. Have a look in here for more details: params.txt
Ground truth and gaze data
IMO ground truth data provided only for the Suburban bridge and Suburban follow sequences:
Suburban_bridge (left and right camera),
Suburban_follow (left and right camera).
Read the following file to know how to interpret the ground truth files: readme.txt
Gaze Data (eye movements of the driver) provided only for the Suburban Bridge sequence:
Suburban-Bridge Gaze data.
Read the following file to know how to interpret the gaze data: readme.pdf
Reference:
Emre Baseski et al., “Road Interpretation for Driver Assistance Based on an Early Cognitive Vision System“, in Proc. VISAPP 2009. (Click here to download this as a bib file).