Skip to content

Latest commit

 

History

History
24 lines (17 loc) · 1.56 KB

di.uoa.gr-DataSet.md

File metadata and controls

24 lines (17 loc) · 1.56 KB
Raw

Our Dataset (Zografou University Campus)

Due to size limitations, this dataset is only available through a set of 3 DVDs. Total Size: 10+GB

Folders 000..005 contain the dataset used to train the networks. Folder calib contains preliminary test data (calibration). Folders 006..008 contain data for testing the system.

In each numbered folder, there is a videos sub-folder, containing left-camera(lc), straight-camera(sc) and right-camera(rc) sub-folders. Each of these contains the videos recorded in .avi format.

  • info.txt contains some description characteristics for each recording trip.
  • script.sh is used for automating the pre-processing of our footage: -- videoParserV3.py splits videos into frames and -- frameSplitterV3.py undistorts image frames and then splits them into three virtual views of 60 degrees

All you need to do is: i. copy videoParserV3.py and frameSplitterV3.py from directory autonomous_drones/tools/camera_rig to each numbered sub-folder ii. copy calibration.2M.yml extracted from widecam_mono_calibration to each numbered sub-folder iii. run from terminal bash ./script.sh in each numbered sub-folder

For calibration, you need to cd in folder calib, then run:

widecam_mono_calibration -input=<path calibration target images> -results=<path to results> OR widecam_mono_calibration -input=./calib.frames.2M/ -results=./calib.frames.2M.results/

The camera calibration application stores the intrinsic parameters in a calibration file (in YML format) and also saves out undistorted images of the calibration target.