Intensity JND comes from Poisson neural noise: Implications for image coding

Research output: Contribution to journalConference articlepeer-review


In audio coding, the loudness of just-noticeable-difference (JND) is proportional to the square root of the loudness Δℓ(ℓ)∝√ℓ(I) below about 5 sones. This is true for both wideband noise and tones with a frequency of 250 Hz or greater. According to J. Allen and J. Neely, the internal noise is Poisson, as would be expected from neutral point process noise. The Weber fraction decreases as one over the square root of the loudness, namely Δℓ/ℓ∝1/√ℓ. Above ℓ = 5 sones, the relative loudness JND Δℓ/ℓ≈0.03. The applicability of this relationship for image coding between brightness B(I) and the brightness JND ΔB(I) may be tested by measuring both the brightness JND and the brightness as a function of intensity, and transforming the intensity JND into a brightness JND, namely ΔB(I) = B(I+ΔI)-B(I)≈ΔI dB/dI.

Original languageEnglish (US)
Pages (from-to)222-233
Number of pages12
JournalProceedings of SPIE - The International Society for Optical Engineering
StatePublished - Jan 1 2000
Externally publishedYes
EventHuman Vision and Electronic Imaging - San Jose, CA, USA
Duration: Jan 24 2000Jan 27 2000

ASJC Scopus subject areas

  • Electronic, Optical and Magnetic Materials
  • Condensed Matter Physics
  • Computer Science Applications
  • Applied Mathematics
  • Electrical and Electronic Engineering

Fingerprint Dive into the research topics of 'Intensity JND comes from Poisson neural noise: Implications for image coding'. Together they form a unique fingerprint.

Cite this