Databases you can use for benchmarking

Pilot
European
Image
Processing
Archive

---

The PCCV Project:
Benchmarking Vision Systems
Overview
Tutorials
Methodology
Case studies
Test datasets
Our image file format
HATE test harness

---

Information
General links
Conferences
Mailing lists
Research groups
Societies

Techniques (CVonline)
Software
Image databases

---

Other stuff
Linux on ThinkPad

Datasets are available in the following categories:

• astronomical imagery
• imagery from books
• CAD models
• content-based image retrieval
• colour
• faces and facial expressions
• forensic
• handwriting
• medical
• motion
• natural scenes
• object recognition
• range data
• remote sensing
• segmentation
• stereo
• surveillance
• texture
• unclassified

Submitting a new entry to the database

We believe this is the most comprehensive listing of image databases anywhere in the world. But if we have somehow managed to omit your database, or the information we have is incorrect, please complete the following form and submit it to us. We'll then amend the page as necessary.

Astronomical Imagery

• database of astronomical test images provided by IAPR technical committee 13

Imagery from Books

• C. A. Glaseby and G. W. Horgan: Image Analysis for the Biological Sciences (John Wiley, 1995)

CAD, 3D Models and Images Thereof

• National Design Repository over 55,000 3D CAD and solid models of (mostly) mechanical/machined engineerign designs. (Format: gif, vrml, wrl, stp, sat) ( Geometric & Intelligent Computing Laboratory / Drexel University)

• UNC's 3D image database many images (Format: GIF)

Content-Based Image Retrieval

• Content-based image retrieval database 11 sets of color images for testing algorithms for content-based retrieval. Most sets have a description file with names of objects in each image. (Format: jpg) ( Efficient Content-based Retrieval Group / University of Washington)

Colour

• Computational Colour Constancy Data A dataset oriented towards computational color constancy, but useful for computer vision in general. It includes synthetic data, camera sensor data, and over 700 images. (Format: tiff) ( Computational Vision Lab / Simon Fraser University)

Faces and Facial Expressions

• AT&T: The Database of Faces (formerly 'The ORL Database of Faces') (Format: PGM)

• Caltech Image Database 450 frontal face images of 27 or so unique people. (Format: GIF)

• CMU PIE Database A database of 41,368 face images of 68 people captured under 13 poses, 43 illuminations conditions, and with 4 different expressions.

• CMU VASC Image Database Images, sequences, stereo pairs (thousands of images) (Format: Sun Rasterimage)

• JAFFE Facial Expression Image Database The JAFFE database consists of 213 images of Japanese female subjects posing 6 basic facial expressions as well as a neutral pose. Ratings on emotion adjectives are also available, free of charge, for research purposes. (Formats: TIFF Grayscale images.)

• The M2VTS multimodal face and head database contains 5 sequences of 37 individuals, comprising imagery and accompanying sound. (Format: homebrew but with conversion instructions)

• MIT face images and more hundreds of images (Format: homebrew)

• The AR Face Database contains over 4,000 color images corresponding to 126 people's faces (70 men and 56 women). Frontal views with variations in facial expressions, illumination, and occlusions. (Format: RAW (RGB 24-bit)) ( Purdue Robot Vision Lab)

• Olivetti face database 400 images (Formats: pgm)

• University of Bern Face images: hundreds of images. (Format: Sun raster)

• University of Oulu Physics-based Face Database: colour images of faces under different illuminants and camera calibration conditions as well as skin spectral reflectance measurements of each person. (Machine Vision and Media Processing Unit, University of Oulu)

• Usenix face database: hundreds of face images. (Format: various)

• Yale Face Database: 165 images of 15 subjects with different lighting, expression, and occlusion configurations.

• Yale Face Database B: 5760 single light source images of 10 subjects each seen under 576 viewing conditions (9 poses x 64 illumination conditions). (Format: PGM) (Center for Computational Vision and Control, Yale University)

Forensic

• FVC2000 Fingerprint Databases FVC2000 is the First International Competition for Fingerprint Verification Algorithms. Four fingerprint databases constitute the FVC2000 benchmark (3520 fingerprints in all). ( Biometric Systems Lab / University of Bologna)

• NIST Fingerprint and handwriting datasets - thousands of images (Format: unknown)

• NIST Fingerprint data compressed multipart uuencoded tar file (Columbia)

Handwriting

• CEDAR CDROM database of handwritten words, ZIP Codes, Digits and Alphabetic characters (Format: unknown)

• NIST Fingerprint and handwriting datasets - thousands of images (Format: unknown)

• UNIPEN database for online handwriting recognition

• US Postal Service (USPS) datasets for The Elements of Statistical Learning

Medical

• El Salvador Atlas of Gastrointestinal VideoEndoscopy Images and Videos of his-res of studies taken from Gastrointestinal Video endoscopy. (Format: jpg, mpg, gif)

• The Mammographic Image Analysis Society (MIAS) mini-database.

• Mammography Image Databases 100 or more images of mammograms with ground truth. Additional images available by request, and links to several other mammography databases are provided. (Format: homebrew)

• NLM HyperDoc Visible Human Project color, CAT and MRI image samples - over 30 images (Format: jpeg)

• University of Massachusetts Vision Image Archive Large image database with aerial, space, stereo, medical images and more. (Format: homebrew)

Motion

• CMU VASC Image Database Images, sequences, stereo pairs (thousands of images) (Format: Sun Rasterimage)

• OSU/SAMPL Database: Range Images, 3D Models, Stills, Motion Sequences over 1000 range images, 3D object models, still images and motion sequences (Format: gif, ppm, vrml, homebrew) ( Signal Analysis and Machine Perception Laboratory / The Ohio State University)

• Optic flow: the Barron and Fleet study.

• Optical flow test image sequences 6 synthetic and 4 real image sequences (Format: Sun Rasterimage)

• Particle image sequences real and synthetic image sequences used for testing a Particle Image Velocimetry application. These images may be used for the test of optical flow and image matching algorithms. (Format: pgm (raw)) ( LIMSI-CNRS/CHM/IMM/vision / LIMSI-CNRS)

• Sequences for optical flow analysis (SOFA) 9 synthetic sequences designed for testing motion analysis applications, including full ground truth of motion and camera parameters. (Format: gif) ( Computer Vision Group / Dept of Computing and Electrical Engineering, Heriot-Watt University)

• University of Massachusetts Vision Image Archive Large image database with aerial, space, stereo, medical images and more. (Format: homebrew)

• USC-SIPI image database including textures, aerial imagery and motion (Format:TIFF)

• VisMouse Database (Format: unknown)

Natural Scenes

• Caltech Image Database 186 images of leaves against different backgrounds. (Format: GIF)

• Groningen Natural Image Database 4000+ 1536x1024 (16 bit) calibrated outdoor images (Format: homebrew)

• IEN Image Library 1000+ images, mostly outdoor sequences (Format: raw, ppm)

• University of Oulu wood and knots database: 1000+ colour images, including classification. (Format: PPM)

• WILD: Weather and Illumination Database WILD is an extensive database of high quality images of an outdoor urban scene, acquired every hour over all seasons. It includes images of the scene taken under a variety of weather and illumination conditions. Each image is tagged with useful ground truth information regarding atmospheric conditions, scene distances and ephemeric data.(Format: unknown)

Object Recognition

• Amsterdam library of object images (ALOI) a collection of colour images of 1000 small objects. There is a systematic variation in viewing angle, illumination angle, and illumination colour for each object, and additionally captured wide-baseline stereo images. Over a hundred images of each object have been captured, yielding a total of 110,250 images for the collection.

• OSU/SAMPL Database: Range Images, 3D Models, Stills, Motion Sequences over 1000 range images, 3D object models, still images and motion sequences (Format: gif, ppm, vrml, homebrew) ( Signal Analysis and Machine Perception Laboratory / The Ohio State University)

• Caltech Image Database about 20 images - mostly top-down views of small objects and toys. (Format: GIF)

• Caltech 360 deg Office Scenes

• COIL-100: 100 everyday objects (Columbia University)

• ETH80 database from ETH Zurich

• View Sphere Database: Images of 8 objects seen from many different viewpoints. The view sphere is sampled using a geodesic with 172 images per sphere. Two sets for training and testing are available. (Format: PPM) (PRIMA, GRAVIR)

• University of Massachusetts Vision Image Archive Large image database with aerial, space, stereo, medical images and more. (Format: homebrew)

• SOIL-47, the Surrey Object Image Library. This is a database of coloured household objects, many of the same shape, viewed over a significant portion of the viewing sphere. The images show mainly multicoloured objects, many of which consist of planar surfaces (boxes) and with generally complex colour structure. The database contains 24 objects with approximately planar surfaces and 22 complex scenes. (Format: PNG)

• Zurich building image database, containing over 1000 images of buildings. (Format: PNG)