Research on Subjective Quality Assessment of Light Field Images

Proceedings of the 2018 VII International Conference on Network, Communication and Computing Pub Date : 2018-12-14 DOI:10.1145/3301326.3301364

Liang Shan, Deyang Liu, P. An, Xinpeng Huang

{"title":"Research on Subjective Quality Assessment of Light Field Images","authors":"Liang Shan, Deyang Liu, P. An, Xinpeng Huang","doi":"10.1145/3301326.3301364","DOIUrl":null,"url":null,"abstract":"Light field (LF) exerts a positive impact on many multimedia applications, due to its variety of the post-processing. For the sheer size of data volume and the huge difference compared to natural 2D images, many processing methods for such particular type of image have been proposed by researchers, for which the image quality assessment (IQA) research for such LF images (LFIs) is urgently needed. However, the development of the IQA of LFIs is restricted for lack of a database that fully reflects characteristics of LFIs. Therefore, in this paper, we build a perceptual quality assessment dataset including 240 distorted images from 8 source images by adding five distortion types with several distortion levels and obtain the mean opinion score (MOS) of each LF image by a developed subjective experiment. Further, we propose a subjective evaluation method to evaluate the perceptual quality of distorted LFIs. In the experiment, some classical full reference IQA metrics were used on the proposed dataset, and the results provide useful insights for the future development of compression solutions and objective IQA for LFIs.","PeriodicalId":294040,"journal":{"name":"Proceedings of the 2018 VII International Conference on Network, Communication and Computing","volume":"1 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2018-12-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"3","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Proceedings of the 2018 VII International Conference on Network, Communication and Computing","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1145/3301326.3301364","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}

引用次数: 3

Abstract

Light field (LF) exerts a positive impact on many multimedia applications, due to its variety of the post-processing. For the sheer size of data volume and the huge difference compared to natural 2D images, many processing methods for such particular type of image have been proposed by researchers, for which the image quality assessment (IQA) research for such LF images (LFIs) is urgently needed. However, the development of the IQA of LFIs is restricted for lack of a database that fully reflects characteristics of LFIs. Therefore, in this paper, we build a perceptual quality assessment dataset including 240 distorted images from 8 source images by adding five distortion types with several distortion levels and obtain the mean opinion score (MOS) of each LF image by a developed subjective experiment. Further, we propose a subjective evaluation method to evaluate the perceptual quality of distorted LFIs. In the experiment, some classical full reference IQA metrics were used on the proposed dataset, and the results provide useful insights for the future development of compression solutions and objective IQA for LFIs.

查看原文本刊更多论文

光场图像主观质量评价研究

光场(LF)由于其后处理的多样性，在许多多媒体应用中发挥着积极的作用。由于数据量庞大，与自然二维图像相比差异巨大，研究者提出了许多针对此类特殊类型图像的处理方法，因此迫切需要对此类LF图像(lfi)进行图像质量评估(IQA)研究。然而，由于缺乏充分反映lfi特征的数据库，制约了lfi的IQA发展。因此，本文通过添加5种不同失真程度的失真类型，构建了包含8幅源图像的240幅失真图像的感知质量评估数据集，并通过开发主观实验获得了每张LF图像的平均意见评分(MOS)。此外，我们还提出了一种主观评价方法来评价扭曲lfi的感知质量。在实验中，一些经典的全参考IQA指标被用于所提出的数据集，结果为lfi的压缩解决方案和客观IQA的未来发展提供了有用的见解。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文求助全文

来源期刊

Proceedings of the 2018 VII International Conference on Network, Communication and Computing

自引率

0.00%

发文量