Biomedical Engineering Reference
In-Depth Information
two-dimensional arrays of sensors. This is because the devices are usually expected to cap-
ture the information at high speed. The geometry of an individual sensor element can be
square, rectangle, or hexagon. Sensitivity to color is achieved by using spectral color filters,
and the spatial or time resolution will be diminished if only a single array of sensors is used
for all colors in a multicomponent color system.
In the proposed system, the imaging array elements have been implemented by enclos-
ing BR solution between two electrodes, or drying a BR film on top of an array of elec-
trodes, or charge-sensitive devices. The element dimensions of the imaging arrays in the
actually implemented imaging arrays are from 20
20 mm
2
down to 2.5
2.5 mm
2
but
even sizes of 15
m
2
have been proposed. In these proposals, the film thickness varies
from 30 nm to about 150
15
8
elements. Even an artificial retina based on BR that mimics the functionality of X-type reti-
nal ganglion cells with motion sensitivity has been proposed (10,17).
In the color-sensitive imaging devices, color sensitivity is based on a combination of BR
and flavin (15), or three types of BR (11,13). The spatial resolution is 3
m, and the spatial resolutions are from 3
2 elements to 8
2 elements (11,13)
or 3
3 elements (15). In the former case, the resolution must be divided by the number
of color components, whereas the latter contains a layered structure to produce a two-
component color response of nine elements. In addition, a model for the color-vision sys-
tem based on BR has been devised (11,19), and it consists of photosensitive and
preprocessing layers.
16.3
What is Color?
Color is among the most salient features of visual environment detected by the human
visual system. Perception of color makes it easier and faster to detect, recognize, discrim-
inate, and classify objects. Moreover, it provides an esthetic component to our visual expe-
riences. Color also gained attention in artificial vision systems and it is central to a number
of image analysis and computer-vision methods.
Color, as commonly understood in hue terms, such as red, green, yellow, etc., does not
have a direct counterpart in a physical world; experience of color is only a product of our
visual perception. The sensation of color originates from responses of three types of light-
sensitive cells in the retina, or more precisely cones, to incoming light. The cone responses
afterwards undergo complex neural processing in the brain, which ultimately results in a
single percept that we call color. Using a physical approach, complete characterization of
color requires a specification of the spectral energy of light originating from the observed
object and approaching the eye (20). Therefore, color is a continuous function of wavelength.
Color is usually represented and communicated using three-dimensional color coordi-
nate systems—tristimulus color spaces, like CIE XYZ, CIELAB, and CIELUV color spaces.
These spaces are closely related to human visual system, in which there are three different
classes of photoreceptors in the retina. A color space can be thought of as a subspace in
infinite-dimensional space of light spectra. The color coordinates associated with the spec-
tra are projections, which are results of inner products between the spectra and basis func-
tions. In the CIE XYZ coordinate system, these functions are color-matching functions. Let
E
(
) be the spectral energy of light originated from the object and
x
i
be a set of basis. The
color coordinates
X
i
,
i
= 1, ...,
n
of the light are calculated as
∫
XEx
i
() ()d
(16.1)
i
