Set 10


SET 10

iROADS Dataset (Intercity Roads and Adverse Driving Conditions)

We release the iROADS dataset as a comprehensive vehicle‐dataset with 4656 image frames, in 7 categories, recorded from moving vehicles on the roads under various weathers and challenging lighting conditions:

CONDITION NO. OF FRAMES
Daylight 903
Night 1050
Rainy day 1049
Rainy night 431
Snowy 569
Sun strokes 307
Tunnel 347
SEQUENCE LENGTH (FRAMES) 4656
COLOUR / GREY Colour
COLOUR DEPTH 24-bit
RESOLUTION (PIXELS) 640 x 360
STEREO RECTIFIED NA
EGO-MOTION INFO NA

The dataset may be used as a unique reference for benchmarking and/or classifiertraining purposes.
Compressed JPG format: iROADS Dataset (ZIP, 89 MB)
None compressed BMP format: iROADS Full Dataset (ZIP, 1.54 GB)

Terms and Conditions

  1. Modification and Commercial Use: The database, in whole or in parts, may not be modified or used for commercial purposes.
  2. Research Purpose: The database is provided to aid researchers in the general areas of computer vision, image processing, and more specifically vehicle detection and driver assistance systems.
  3. Indemnification: The researcher who use this dataset agree that the collected dataset‐iROADS is solely provided for research purposes and the providers will not be responsible for any kind of losses, expenses, damages, demands and/or claims based upon usage of the provided dataset.
SEQUENCE NAME TOTAL FRAMES
Daylight 903
SEQUENCE NAME TOTAL FRAMES
Night 1050
SEQUENCE NAME TOTAL FRAMES
Rainy day 1049
SEQUENCE NAME TOTAL FRAMES
Rainy night 431
SEQUENCE NAME TOTAL FRAMES
Snowy 569
SEQUENCE NAME TOTAL FRAMES
Sun strokes 307
SEQUENCE NAME TOTAL FRAMES
Tunnel 347

Reference:

Mahdi Rezaei and Mutsuhiro Terauchi, “Vehicle Detection Based on Multi-feature Clues and Dempster-Shafer Fusion Theory“, Pacific-Rim Symposium on Image and Video Technology (PSIVT), LNCS, pp. 60-72, 2014. (Click here to download this as a bib file).