| Contents | Previous | Next |
Color imaging is one of the fundamental components of any graphics system, and it is often a source of great complexity in the imaging model. The Java 2D™ API provides support for high-quality color output that is easy to use and allows advanced clients to make sophisticated use of color.
The key color management classes in the Java 2D
API are ColorSpace, Color, ColorModel:
ColorSpace represents a system for
measuring colors, typically using three separate numerical values
or components. The ColorSpace class
contains methods for converting between the color space and two
standard color spaces, CIEXYZ and
RGB.Color is a fixed color, defined in
terms of its components in a particular ColorSpace. To draw a Shape in a color, such as red, you pass a
Color object representing that color to
the Graphics2D context. Color is defined in the java.awt
package.ColorModel describes a particular
way that pixel values are mapped to colors. A ColorModel is typically associated with an
Image or BufferedImage and provides the information necessary
to correctly interpret the pixel values. ColorModel is defined in the java.awt.image package.A ColorModel is used to
interpret pixel data in an image. This includes mapping components
in the bands of an image to components of a particular color space.
It might also involve extracting pixel components from packed pixel
data, retrieving multiple components from a single band using
masks, and converting pixel data through a lookup table.
To determine the color value of a particular pixel
in an image, you need to know how color information is encoded in
each pixel. The ColorModel associated
with an image encapsulates the data and methods necessary for
translating a pixel value to and from its constituent color
components.
The Java 2D™ API provides two color models
in addition to the DirectColorModel and
IndexColorModel defined in the JDK 1.1
software release:
ComponentColorModel can handle an
arbitrary ColorSpace and an array of
color components to match the ColorSpace. This model can be used to represent most
color models on most types of GraphicsDevices.PackedColorModel is a base class for
models that represent pixel values that have their color components
embedded directly in the bits of an integer pixel. A PackedColorModel stores the packing information that
describes how color and alpha components are extracted from the
channel.The DirectColorModel in the JDK
1.1 software release is a PackedColorModel.A ColorSpace object
represents a system for measuring colors, typically using three
separate numeric values. For example, RGB and CMYK are color
spaces. A ColorSpace object serves as a
colorspace tag that identifies the specific color space of a
Color object or, through a ColorModel object, of an Image, BufferedImage, or
GraphicsConfiguration. ColorSpace provides methods that transform
Colors in a specific color space to and
from sRGB and to and from a well-defined
CIEXYZ color space.
All ColorSpace objects
must be able to map a color from the represented color space into
sRGB and transform an sRGB color into the represented color space. Since
every Color contains a ColorSpace object, set explicitly or by default,
every Color can also be converted to
sRGB. Every GraphicsConfiguration is associated with a
ColorSpace object that in turn has an
associated ColorSpace. A color specified
in any color space can be displayed by any device by mapping it
through sRGB as an intermediate color
space.
The methods used for this process are toRGB and fromRGB:
toRGB transforms a Color in the represented color space to a
Color in sRGB.fromRGB takes a Color in sRGB and
transforms it into the represented color space.Though mapping through sRGB always works, it's not always the best
solution. For one thing, sRGB cannot
represent every color in the full gamut of CIEXYZ colors. If a color is specified in some space
that has a different gamut (spectrum of representable colors) than
sRGB, then using sRGB as an intermediate space results in a loss of
information. To address this problem, the ColorSpace class can map colors to and from another
color space, the “conversion space” CIEXYZ.
The methods toCIEXYZ
and fromCIEXYZ map color values from the
represented color space to the conversion space. These methods
support conversions between any two color spaces at a reasonably
high degree of accuracy, one Color at a
time. However, it is expected that Java 2D API implementations will
support high-performance conversion based on underlying platform
color-management systems, operating on entire images. (See
ColorConvertOp in “Imaging” on
page 67.)
Figure 6-1 and Figure 6-2 illustrate the process
of translating a color specified in a CMYK color space for display
on an RGB color monitor. Figure 6-1 shows a mapping
through sRGB. As this figure
illustrates, the translation of the CMYK color to an RGB color is
not exact because of a gamut mismatch. 1
Figure 6-2 shows the same process
using CIEXYZ as the conversion space.
When CIEXYZ is used, the color is passed
through accurately.
ColorSpace is actually
an abstract class. The Java 2D API provides one implementation,
ICC_ColorSpace, which is based on ICC
Profile data as represented by the ICC_Profile class. You can define your own
subclasses to represent arbitrary color spaces, as long as the
methods discussed above are implemented. However, most developers
can simply use the default sRGB
ColorSpace or color spaces that are
represented by commonly available ICC Profiles, such as profiles
for monitors and printers, or profiles embedded in image data.
“ColorSpace” on
page 90 describes how ColorSpace objects represent a color space and how
colors in the represented space can be mapped to and from a
conversion space. Color management systems are often used to handle
the mapping between color spaces. A typical color management system
(CMS) manages ICC profiles, which are similar to ColorSpace objects; ICC profiles describe an input
space and a connection space, and define how to map between them.
Color management systems are very good at figuring out how to map a
color tagged with one profile into the color space of another
profile.
The Java 2D API defines a class called
ICC_Profile that holds data for an
arbitrary ICC Profile. ICC_ColorSpace is
an implementation of the abstract ColorSpace class. ICC_ColorSpace objects can be constructed from
ICC_Profiles. (There are some
limitations—not all ICC Profiles are appropriate for defining
an ICC_ColorSpace).
ICC_Profile has several
subclasses that correspond to specific color space types, such as
ICC_ProfileRGB and ICC_ProfileGray. Each subclass of ICC_Profile has a well-defined input space (such as
an RGB space) and a well-defined connection space (like
CIEXYZ). The Java 2D API can use a
platform's CMS to access color profiles for various devices
such as scanners, printers, and monitors. It can also use the CMS
to find the best mapping between profiles.
The Color class
provides a description of a color in a particular color space. An
instance of Color contains the value of
the color components and a ColorSpace
object. Because a ColorSpace object can
be specified in addition to the color components when a new
instance of Color is created, the
Color class can handle colors in any
color space.
The Color class has a
number of methods that support a proposed standard RGB color space
called sRGB (see http://www.w3.org/pub/WWW/Graphics/Color/sRGB.html).
sRGB is the default color space for the
Java 2D API. Several constructors defined by the Color class omit
the ColorSpace parameter. These
constructors assume that the color's RGB values are defined in
sRGB, and use a default instance of
ColorSpace to represent that space.
The Java 2D API uses sRGB as a convenience to application programmers,
not as a reference color space for color conversion. Many
applications are primarily concerned with RGB images and monitors,
and defining a standard RGB color space makes writing such
applications easier. The ColorSpace
class defines the methods toRGB and
fromRGB so that developers can easily
retrieve colors in this standard space. These methods are not
intended to be used for highly accurate color correction or
conversions. See “ColorSpace” on
page 90 for more information.
To create a color in a color space other than
sRGB, you use the Color constructor that takes a ColorSpace object and an array of floats that
represent the color components appropriate to that space. The
ColorSpace object identifies the color
space.
To display a rectangle of a certain color, such as the process color cyan, you need a way to describe this color to the system. There are a number of different ways to describe a color; for example, a color could be described as a set of red, green, and blue (RGB) components, or a set of cyan, magenta, yellow, and black (CMYK) components. These different techniques for specifying colors are called color spaces.
As you probably know, colors on a computer screen are generated by blending different amounts of red, green, and blue light. Therefore, using an RGB color space is standard for imaging on computer monitors. Similarly, four-color process printing uses cyan, magenta, yellow, and black ink to produce color on a printed page; the printed colors are specified as percentages in a CMYK color space.
Due to the prevalence of computer monitors and color printing, RGB and CMYK color spaces are both commonly used to describe colors. However, both types of color spaces have a fundamental drawback—they are device-dependent. The cyan ink used by one printer might not exactly match the cyan ink used by another. Similarly, a color described as an RGB color might look blue on one monitor and purplish on another.
The Java 2D API refers to RGB and CMYK as color space types. A particular model of monitor with its particular phosphors defines its own RGB color space. Similarly, a particular model of printer has its own CMYK color space. Different RGB or CMYK color spaces can be related to each other through a device-independent color space.
Standards for the device-independent specification
of color have been defined by the International Commission on
Illumination (CIE). The most commonly used device-independent color
space is the three-component XYZ color space developed by CIE. When
you specify a color using CIEXYZ, you
are insulated from device dependencies.
Unfortunately, it’s not always practical to
describe colors in the CIEXYZ color
space—there are valid reasons for representing colors in
other color spaces. To obtain consistent results when a color is
represented using a device-dependent color space such as a
particular RGB space, it is necessary to show how that
RGB space relates to a device-independent space like
CIEXYZ.
One way to map between color spaces is to attach
information to the spaces that describes how the device-dependent
space relates to the device-independent space. This additional
information is called a profile. A
commonly used type of color profile is the ICC Color Profile, as
defined by the International Color Consortium. For details, see the
ICC Profile Format Specification, version 3.4 available at
http://www.color.org.
Figure 6-3 illustrates how a solid color and a scanned image are passed to the Java 2D API, and how they are displayed by various output devices. As you can see in Figure 6-3, both the input color and the image have profiles attached.
Once the API has an accurately specified color, it must reproduce that color on an output device, such as a monitor or printer. These devices have imaging characteristics of their own that must be taken into account to make sure that they produce the correct results. Another profile is associated with each output device to describe how the colors need to be transformed to produce accurate results.
Achieving consistent and accurate color requires that both input colors and output devices be profiled against a standard color space. For example, an input color could be mapped from its original color space into a standard device-independent space, and then mapped from that space to the output device’s color space. In many respects, the transformation of colors mimics the transformation of graphical objects in an (x, y) coordinate space. In both cases, a transformation is used to specify coordinates in a “standard” space and then map those coordinates to a device-specific space for output.
| Contents | Previous | Next |
