Generalized image contrast enhancement
technique based on the Heinemann contrast
Calvin F. Nodine
University of Pennsylvania
Department of Radiology
308 Stemmler Hall
36th and Hamilton Walk
Philadelphia, Pennsylvania 19104-6086
This paper presents a generalized image contrast en-
hancement technique, which equalizes the perceived brightness
distribution based on the Heinemann contrast discrimination model.
It is based on the mathematically proven existence of a unique so-
lution to a nonlinear equation, and is formulated with easily tunable
parameters. The model uses a two-step log-log representation of
luminance contrast between targets and surround in a luminous
background setting. The algorithm consists of two nonlinear gray
scale mapping functions that have seven parameters, two of which
are adjustable Heinemann constants. Another parameter is the
background gray level. The remaining four parameters are nonlinear
functions of the gray-level distribution of the given image, and can
be uniquely determined once the previous three are set. Tests have
been carried out to demonstrate the effectiveness of the algorithm
for increasing the overall contrast of radiology images. The tradi-
tional histogram equalization can be reinterpreted as an image en-
hancement technique based on the knowledge of human contrast
perception. In fact, it is a special case of the proposed algorithm.
© 1996 SPIE and IS&T.
Contrast enhancement techniques are widely used to im-
prove the visibility of the structures of interest in images
presented to human viewers. Therefore, it is quite natural to
take into account human contrast perception characteristics
in image enhancement. There have been two approaches
for incorporating contrast perception models into digital
image enhancement. Johnston et al.,
Blume et al.,
others proposed a perceptual linearization concept for CRT
displays. The goal is to modify the mapping from digital
gray levels to displayed luminance so that equal increments
in digital levels result in equal number of just-noticeable
difference ͑JND͒ steps in contrast perception. This image-
independent technique enables standardized displays across
all CRT monitors by using a ﬁxed lookup table tailored to
the monitor to transform all images in the same way.
A more complex approach is to modify the gray-level
distribution of the given image to achieve uniformly dis-
tributed perceived brightness levels. Perceived brightness,
is the subjective attribute of any light sensa-
tion giving rise to the perception of luminous intensity. For
the purpose of quantitative analysis, brightness is com-
monly expressed using a rating scale based on a decision
variable such as a JND,
or contrast perception threshold.
This approach combines human contrast perception models
with information theory, which states that a uniformly dis-
tributed signal has the maximum entropy. When all per-
ceived brightness levels of the processed image are made
equiprobable, the contrast aspect of visual information de-
livery is believed to be maximized. Ordinary histogram
equalization is an attempt in this direction without taking
into account a speciﬁc contrast perception model. Unlike
the ﬁrst approach, these techniques provide image-
dependent lookup tables, or mappings, that are called to
equalize perceived brightness.
There are many models that depict the relationship be-
tween contrast perception thresholds and displayed lumi-
nance variations. Among them are Weber’s logarithmic
law, Steven’s power law, and the background ratio,
the ﬁrst model being most widely used. Frei
histogram hyperbolization algorithm, which equalizes per-
ceived brightness assuming that JNDs are described by We-
ber’s law. Heinemann
and many others have pointed out
that Weber’s law does not hold at low and high intensity
ranges. Heinemann models contrast perception as the si-
multaneous discrimination of a target in a local luminous
environment embedded in a luminous background ͓Fig.
1͑b͔͒. Unlike Weber’s experiment ͓Fig. 1͑a͔͒, which is es-
sentially a signal detection problem, Heinemann’s experi-
Current afﬁliation: National Institute of Health, Bethesda, Maryland.
Paper 94-041 received Dec. 28, 1994; revised manuscript received Apr. 1, 1996;
accepted for publication Apr. 1, 1996.
1017-9909/96/$6.00 © 1996 SPIE and IS&T.
Journal of Electronic Imaging 5(3), 388
395 (July 1996).
388 / Journal of Electronic Imaging / July 1996 / Vol. 5(3)