A New Framework for Measuring 2D and 3D Visual Information in Terms of Entropy

Kwanghyun Lee, Sanghoon Lee

Research output: Contribution to journalArticlepeer-review

13 Citations (Scopus)


In recent years, the problem of how to more perceptually quantify visualizations of an object and a surface displayed in 3D space between the human eye and a display has been more prominent with rapid increase in the demand for 3D/ultrahigh-definition content. In order to quantify the content information in terms of human visual perception, it is necessary to measure the visual information of 2D and 3D videos accurately in accordance with the human visual system. In this paper, we investigate a new framework for expressing visual information in bits termed visual entropy, based on information theory. 2D visual entropy (2DVE) is composed of two major components: 1) texture entropy on the 2D surface and 2) depth entropy based on the monocular cue. In contrast, 3D visual entropy (3DVE) includes the depth entropy based on the binocular cue in addition to the 2DVE. A series of simulations is conducted to demonstrate the effectiveness of visual entropy, including the degree of accuracy needed to represent perceptual information by means of the texture and monocular and binocular depth entropies. In the simulation results, two successful example applications are also provided, whereby visual entropy is applied to the problems of predicting the visual discomfort experienced when viewing 3D displays and of analyzing performance tradeoffs between 2D and 3D contents.

Original languageEnglish
Pages (from-to)2015-2027
Number of pages13
JournalIEEE Transactions on Circuits and Systems for Video Technology
Issue number11
Publication statusPublished - 2016 Nov

Bibliographical note

Publisher Copyright:
© 2015 IEEE.

All Science Journal Classification (ASJC) codes

  • Media Technology
  • Electrical and Electronic Engineering


Dive into the research topics of 'A New Framework for Measuring 2D and 3D Visual Information in Terms of Entropy'. Together they form a unique fingerprint.

Cite this