The Vault

Improving the Efficiency of Video Coding by using Perceptual Preprocessing Filter
Research Paper / Jan 2012

Improving the Efficiency of Video Coding by using
Perceptual Preprocessing Filter


Rahul Vanam and Yuriy A. Reznik,

InterDigital Communications, LLC, 9710 Scranton Road, San Diego, CA 92121 USA
E-mail:{rahul.vanam, yuriy.reznik}@interdigital.com

We propose design of a pre-processing filter that improves effectiveness of video coding
by removing spatial oscillations that are invisible under given viewing conditions, such as
display contrast, pixel density, and a distance between user and the screen. Figure 1
shows an exemplary deployment of a mobile video delivery system using such a filter.

DecoderEncoder

Perceptual
preprocessing

filter

Sensors

Display

Viewing distance, display
density, contrast ratio

Network


Figure 1: Video delivery system utilizing perceptual pre-processing filter.

The block-diagram of our proposed filter is shown in Figure 2.

Input video/
image

Output
video/image

-+
-

Viewing
distance

Display pixel
density

Contrast ratio

Convert to
linear color

space

Extract
Luminance

Black-level
adjustment

Compute
local DCs:

[DC]xy

Compute
amplitudes:

[A]xy

Estimate local
contrasts:

[C]xy = [A]xy /
[DC]xy

Estimate cut-off
frequencies:

[Fc]xy = F(Cxy)

Convert to
original color

space

Adaptive low-
pass filter


Figure 2: Block diagram of our perceptual filter.

The underlying phenomenon exploited by our filter is known as Contrast Sensitivity
Function (CSF) of human vision [1]. We use Movshon and Kiorpes CSF model [2].
In order to apply it, we extract luminance channel, adjust black level to account contrast
of the screen, and then compute local averages and amplitudes of oscillations in the span
of about 2o of viewing angle. We then compute local contrasts of oscillations and then use the
CSF model to estimate highest visible frequency in such regions. We use such frequencies to
control a low-pass filter applied to an image.

In our experiments with 4.3”, 720p-resolution screens, held at distances in 12–24” range,
and ambient light of 0–2000 lux, we have observed savings of up to 70% of bitrate.
We have used H.264 video codec and standard video test sequences in our tests.

References
[1] A. C. Bovik, “Handbook of image and video processing”. Academic Press, 2005.
[2] J. Movshon and L. Kiorpes, “Analysis of the development of spatial contrast sensitivity in monkey and

human infants,” JOSA A, vol. 5, no. 12, pp. 2166–2172, 1988.