Important Parameters for Image Color Analysis: An Overview - Visual Computing: Scientific Visualization and Imaging Systems - page 84

conversion of RGB values in other systems, such as HSL, HSI, HSB, and HSV that
are widely employed since they consider concepts of hue, saturation, and

brightness.
For calibration of digital cameras it is necessary to convert the different color
systems used into a color system that is known to be subsequently compared with
standards previously calibrated in order to give the system the necessary trace-
ability. The color system used in such calibrations is the CIEXYZ color system,
developed by the CIE to be used in colorimetry and, from this, other color systems
can be used according to the need of the application area. In the industrial area the
most used color system is the CIELAB because it is the closest to human
perception.

4.4.1 Parameters Influence

Because an object's color is one of the main indicators of perception of the
products quality, the factors that influence color analysis should be considered in
the system of visual assessment.
The image acquisition step is a physical process that depends on a large number
of parameters for obtaining a quality final image. Some parameters influence the
image quality to be obtained, such as the device used to acquire the image, the
lighting system used, and the background color. There are many types of devices
used in image acquisition, such as digital cameras, scanners, video cameras, and
thermal. The most commonly used devices are digital cameras, which cover about
58 % of the total of devices [ 1 ]. Another important parameter in the image
acquisition device is a detection system, which can be CCD or CMOS, since each
has differences in the speed of image acquisition and sensitivity to light, both of
which have advantages and disadvantages that must be studied case by case.
The type of lighting is also of influence in the resulting picture. It is important
to note that some features of the light source used, such as CCT and color ren-
dering index (CRI), are decisive in the final perception of the image and should be
set according to the research objective [ 5 ]. There are many light sources available,
which have quite different properties between each other. Thus, to standardize the
identification of colors it is necessary to define some characteristics for the light
sources, such as the CRI and CCT [ 4 ]. With the knowledge of the Spectral Power
Distribution (SPD) it is possible to calculate the CCT that expresses the color
appearance of the light emitted by the source. The CCT increases from red to blue;
the higher its value is, the brighter will be the color hue of the light, and the lower,
the more yellowish, thus influencing in the perception of the color of the material.
Lamps with CCT below 3,100 K are considered ''hot'', and those with values
above 4,100 K are considered ''cold'' while those in the range of 3,100-4,100 K,
are considered neutral color or moderate (Table 4.1 ).
The CRI is also an important parameter, quantifying the fidelity with which the
colors are reproduced when struck by a particular source [ 6 ]. The CRI is calculated