Windowed 5 DOF Light Field Image Database (Win5-LID)

Introduction

Recently, light field image has attracted wide attention. However, much less work has been conducted on the perceptual evaluation of light field image. In this work, we create the first windowed 5 degree of freedom light field image database (Win5-LID) based on stereoscopic display, which provides windowed 5 DOF experience and all the depth cues of light field image. The database consists of light field images with representative compression and reconstruction artifacts.

Figure 1. Double-Stimulus Continuous Quality Scale (DSCQS)

Figure 2. 6 real scenes (SRC01-SRC06) and 4 synthetic scenes (SRC07-SRC10)

Figure 3. Distortion Type and Database Information

>

Video 1. Windowed 5 DOF Light Field Image Display Mode Demo

Database Description

Win5-LID contains 6 real scenes captured by Lytro illum and 4 synthetic scenes as pristine images. The selected contents cover a wide range of key low attributes and contain abundant semantic features, such as human, plant and object. All content are of identical angular resolution 9*9, and the spatial resolutions of real scenes and synthetic scenes are 434*625 and 512*512, respectively. Win5-LID consist of two types of representative artifacts, the compression and reconstruction distortions. Here compression schemes use JPEG2000 and HEVC encoders, and reconstruction algorithms (i.e. linear interpolation (LN), nearest neighbor interpolation (NN), and two CNN model are used to process image. Finally, Win5-LID contains 220 distortion light field images. In the Win5-LID, it provide all depth cues of light field image and extend double-stimulus continuous quality scale (DSCQS) method with 5 discrete scales. And the associated picture mean opinion score (MOS) and overall MOS value is provided for each light field image.

Get Database

We are making Stereoscopic Image Retargeting Database available to the research community free of charge. If you use this database in your research, we kindly ask that you reference our paper listed below:

[1] Shi L, Zhao S, Zhou W, et al. Perceptual Evaluation of Light Field Image[C]//2018 25th IEEE International Conference on Image Processing (ICIP). IEEE, 2018: 41-45.

To get the publicly available release of the database by clicking THIS link. Please fill THIS FORM and the source code will be sent to you.

Copyright Notice

Copyright (c) 2018 University of Science and Technology of China

All rights reserved.

Permission is hereby granted, without written agreement and without license or royalty fees, to use, copy, modify, and distribute this database and its documentation for any purpose, provided that the copyright notice in its entirety appear in all copies of this database, and the original source of this database, Immersive Media Computing Lab (IMCL) at University of Science and Technology of China (USTC), is acknowledged in any publication that reports research using this database. The research is to be cited in the bibliography whenever the database is used as:

Likun Shi, Shengyang Zhao, Wei Zhou and Zhibo Chen "Perceptual Evaluation of Light Field Image".

IN NO EVENT SHALL THE UNIVERSITY OF SCIENCE AND TECHNOLOGY OF CHINA BE LIABLE TO ANY PARTY FOR DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES ARISING OUT OF THE USE OF THIS DATABASE AND ITS DOCUMENTATION, EVEN IF UNIVERSITY OF SCIENCE AND TECHNOLOGY OF CHINA HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.

THE UNIVERSITY OF SCIENCE AND TECHNOLOGY OF CHINA SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE. THE DATABASE PROVIDED HEREUNDER IS ON AN "AS IS" BASIS, AND UNIVERSITY OF SCIENCE AND TECHNOLOGY OF CHINA HAS NO OBLIGATION TO PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.