Start page

VCL@FER image quality assessment database

Introduction

Assessment of image quality is an open problem today. Subjective image quality assessment (IQA) is based on subjective experiments in which image quality has been evaluated by human observers perceiving and ranking images.

The results of such experiments depend on psychological processes of perception. Though reliable, because it depends on psycho-visual perception of the each individual that is assessing image quality, subjective method is expensive, difficult to design and time consuming to compute. Several critical factors of the human observers can influence on the final results of assessment such as environmental conditions, motivation and mood of the observers. On the other hand, subjective IQA allows a better understanding of the mechanisms underlying quality perception, providing useful information for the subsequent modeling phase.

VCL@FER database

VCL@FER database consists of 575 images, 23 out of which are originals, with 4 different degradation types per image and 6 degrees of degradation.

The degradation types are:

Testing was conducted on 118 subjects who had to grade 11307 images in total (around 96 images per subject). Every image was evaluated between 16 and 36 times. Every subject had to grade different types of degradation in one session, e.g. sessions were not divided on degradation types. This will influence on objective measure performances. Mean Opinion Score (MOS) was calculated according to the Rec. ITU-R BT.500-11.

Original images


VCL@FER database release agreement

You can download the VCL@FER database here under the following terms:

  1. The VCL@FER database will not be further distributed, published, copied, or further disseminated in any way or form whatsoever, whether for profit or not. This includes further distributing, copying or disseminating to a facility or organization unit in the requesting university, organization, or company. VCL@FER images may be used to create new images but the downloaded images will not be modified.
  2. All the images will be used for the purpose of academic or scientific research only. The VCL@FER database, in whole or in part, will not be used for any commercial purpose in any form. Commercial distribution or any act related to commercial use of this database is strictly prohibited.
  3. All documents and papers that report research results obtained using the VCL@FER database will acknowledge the use of the VCL@FER database. Please cite the following paper:
    A.Zaric, N.Tatalovic, N.Brajkovic, H.Hlevnjak, M.Loncaric, E.Dumic, S.Grgic, "VCL@FER Image Quality Assessment Database", AUTOMATIKA Vol. 53, No. 4, pp. 344–354, 2012
  4. While every effort has been made to ensure accuracy, VCL@FER database owners cannot accept responsibility for errors or omissions.
  5. Use of VCL@FER database is free of charge.
  6. VCL@FER database owners reserve the right to revise, amend, alter or delete the information provided herein at any time, but shall not be responsible for or liable in respect of any such revisions, amendments, alterations or deletions.

Results

Download images

MOS results