The .enpeda.. Image Sequence Analysis Test Site (EISATS) offers sets of image sequences for the purpose of comparative performance evaluation of stereo vision, optic flow, motion analysis, or further techniques in computer vision.

Information on Data

All grey-scale images are in PGM (Portable Grey Map) format and all colour images are in PPM (Portable Pixel Map) format. All binary data is Big Endian (as the current standard). Image depths are either 8-bit or 12-bit. (i.e., PGM has 1 channel of 8/12-bit, PPM has 3 channels of 8/12-bit).These can be viewed through good image viewers (such as the freeware Infranview, which can also batch convert to other image formats) and can be read by OpenCV (C++) and Matlab. All data is compressed using 7-Zip format.
How to cite our datasets:
We grant permission to use and publish all images and maps on this EISATS website. However, if you use our datasets, we request that you cite the appropriate paper:

  • [1] for data of Set 1 or Set 2 (sequences 1 and 2).
  • [2] for data of Set 2 (sequence 1-ECCV).
  • [3] for data of Set 3.
  • [4] for data of Set 4.
  • [5] for data of Set 5.
  • [6] for data of Set 6 (sequences: Parking Lot 1, 2 and 3).
  • [7] for data of Set 6 (sequences: Containers, Trailer and Wall).
  • [8] for data of Set 7.
  • [9] for data of Set 8.
  • [10] for data of Set 9 (sequences: Barriers, Harbour bridge, Queen street and People).
  • [11] for data of Set 9 (sequences: Dusk, Midday, Night and Wiper).
  • [12] for data of Set 10.
  • [13] for data of Set 11.

For FAQ’s, click here.

SET 1: Night vision stereo sequences (Daimler AG)

These seven stereo night vision sequences (12 bit, between 220 and 300 pairs of frames each) have been provided by Daimler AG, Germany, in June 2007 (group of Dr. Uwe Franke). These sequences come with ego-motion data and time stamps for each frame.

SET 2: Synthesized stereo sequences (.enpeda.. & Daimler AG)

These synthesized stereo sequences (with ground truth) have been provided by Tobi Vaudrey (.enpeda..) and Clemens Rabe (Daimler AG).

SET 3: IMO’s in color stereo sequences (Drivsco)

These three day-time, color stereo sequences have been provided by the European Drivsco project. Independent moving objects ground truth and gaze data is now available.

SET 4: Grey-level binocular stereo sequences (Hella Aglaia Mobile Vision & .enpeda..)

These day-time, gray-level stereo sequences have been provided by Hella Aglaia Mobile Vision GmbH, Germany.

SET 5: 8-bit grey-level trinocular stereo sequences (.enpeda..)

Trinocular stereo sequences (rectified by pairs) captured with HAKA1.

SET 6: Grey-level stereo sequences with range scans (HU Berlin and .enpeda../Daimler A.G.)

Stereo vision sequences where the test vehicle drives thru a car park, ground truth from laser scanner, sgm, block and cross matcher disparity maps included.

SET 7: Grey-level stereo for scene labeling analysis (Daimler AG)

Stereo sequences with ground truth for scene labeling analysis (segmentation).

SET 8: Colour monocular sequences (.enpeda..)

Monocular sequences recorded at high frame rates for optic flow detection.

SET 9: 10-bit grey-level trinocular sequences (.enpeda..)

10-bit trinocular sequences captured with HAKA1.

SET 10: iROADS (Intercity Roads and Adverse Driving Conditions)

24-bit binocular sequences in 7 categories, under various weather and lighting conditions.

SET 11: Monocular driver monitoring under different lighting conditions

These 6 video sequences were recorded for different drivers and can be used for evaluating eye-status detection.


[1] 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).

[2] Andreas Wedel, Clemens Rabe, Tobi Vaudrey, Thomas Brox, Uwe Franke and Daniel Cremers, “Efficient Dense Scene Flow from Sparse or Dense Stereo Data“, in Proc. 10th European Conference on Computer Vision (ECCV ’08), 739-751, 2008. (Click here to download this as a bib file).

[3] 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).

[4] Reinhard Klette, Norbert Krüger, Tobi Vaudrey, Karl Pauwels, Marc van Hulle, Sandino Morales, Farid Kandil, Ralf Haeusler, Nicolas Pugeault, Clemens Rabe, and Markus Lappe, “Performance of Correspondence Algorithms in Vision-Based Driver Assistance using an online Image Sequence Database“, IEEE Trans. Vehicular Technonlogy, to appear, 2011.

[5] Sandino Morales and Reinhard Klette, “A third eye for performance evaluation in stereo sequence analysis“, in Proc. 13th International Conference on Computer Analysis of Images and Patterns (CAIP ’09), 1078-1086, 2009. (Click here to download this as a bib file).

[6] Reulke, Ralf und Luber, Andreas und Haberjahn, Mathias und Piltz, Björn (2009) “Validierung von mobilen Stereokamerasystemen in einem 3D-Testfeld“. 3D-NordOst , 10.12.-11. Dec. 2009 , Berlin.

[7] Sandino Morales and Reinhard Klette, “Ground Truth Evaluation of Stereo Algorithms for Real World Applications“, In Proc. Computer Vision ACCV 2010 Workshops, 152-162, 2010. (Click here to download as a bib file).

[8] Alexander Barth, Jan Siegemund, Annemarie Meissner, Uwe Franke and Wolfgang Förstner, “Probabilistic Multi-Class Scene Flow Segmentation for Traffic Scenes“, in Proc. DAGM 2010. (Click here to download this as a bib file).

[9] Kun Ju, Simon Hermann and Reinhard Klette, “Optic Flow for Slow Motion“, Tech. Rep. 68, The .enpeda.. Project, The University of Auckland, 2011. (Click here to download this as a bib file).

[10] Simon Hermann, Sandino Morales, and Reinhard Klette, “Half Resolution Semi-Global Matching“, In Proc.2011 IEEE Intelligent Vehicles Symposium (IEEE IV ’11) (Click here to download this as a bib file).

[11] Konstantin Schauwecker, Sandino Morales, Simon Hermann and Reinhard Klette, “A Comparative Study of Stereo-Matching Algorithms for Road Modelling in the Presence of Windscreen Wipers“, In Proc. 2011 IEEE Intelligent Vehicles Symposium (IEEE IV ’11) (Click here to download this as a bib file).

[12] Mahdi Rezaei and Mutsuhiru Terauchi, “Vehicle Detection Based on Multi-feature Clues and Dempster-Shafer Fusion Theory“, In Proc. 6th Pacific-Rim Symposium on Image and Video Technology (PSIVT ’13, LNCS, ’14) (Click here to download this as a bib file).

[13] Michal Daniluk, Mahdi Rezaei, Radu Nicolescu, and Reinhard Klette. Eye status based on eyelid detection: A driver assistance system. In Proc. Int. Conf. Computer Vision Graphics (ICCVG), LNCS, Springer, 2014.

Skip to toolbar