Television – Format conversion – Specified chrominance processing
Reexamination Certificate
1999-08-25
2001-04-03
Lee, Michael (Department: 2614)
Television
Format conversion
Specified chrominance processing
C348S659000, C345S156000
Reexamination Certificate
active
06211917
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The invention relates to a method and apparatus for performing color space conversion, more particularly to a method and apparatus requiring a fewer number of look-up tables for converting digitized luminance-chrominance color space signals to digitized RGE color space signals.
2. Description of the Related Art
It is desirable to merge a video signal with graphic signals in a multi-media computer system. The video signal may come from a television image processing system having a capture or frame grabbing capability, or from a compressed video playback of a CD-ROM or network transmission. Color space conversion is needed in image processing applications to convert luminance-chrominance color space signals, which offer the advantages of a lower transmission bandwidth and a lower data storage requirement, into RGB color space signals, which are used when displaying an image on a computer monitor.
CCIR 601, which was proposed by the Comite Consultatif International des Radiocommunications (CCIR), establishes the following formulas for converting from the YCbCr luminance-chrominance color space to the RGB color space:
R=Y+
1.402(
Cr-
128) (a.1)
G=Y−
0.714(
Cr-
128)−0.344(
Cb-
128) (a.2)
B=Y+
1.772(
Cb−
128) (a.3)
If U and V are used to represent the shifted chrominance components (Cb-128) and (Cr-128), respectively, Equations (a.1) to (a.3) can be rewritten as follows:
R=Y+
1.402
V
(b.1)
G=Y−
0.714
V−
0.344
U
(b.2)
B=Y+
1.772
U
(b.3)
where Y ranges between [0, 2551,] and U and V range between [−128, 127] in an 8-bit representation for each of the Y, Cb and Cr color space components.
Color space conversion is often implemented by employing multipliers or look-up tables to achieve the matrix multiplication operations. Look-up tables are preferred because of their less complicated constructions. It is noted that the matrix multiplication operations dominate the hardware complexity of a color space converting apparatus. As such, the number of look-up tables is critical in determining the cost of implementing the color space converting apparatus. To implement the YCbCr to RGB color space conversion of Equations (a.1) to (a.3), a conventional color space converter usually requires four look-up tables to perform the matrix multiplication of chrominance components. Although the use of four look-up tables is less expensive to implement as compared to another conventional color space converter which uses a 3-by-3 multiplication matrix, a further reduction in the number of look-up tables is desirable.
SUMMARY OF THE INVENTION
Therefore, the object of the present invention is to provide a method and apparatus requiring less than four look-up tables for converting digitized luminance-chrominance color space signals to digitized RGB color space signals.
Accordingly, it is found that Equations (b.1) to (b.3) can be rearranged as follows to result in RGB color combination signals by linearly combining the conversion formulas:
R−G=
0.714(2
V
)+0.344(
U+
2
V
) (c.1)
B−G=
0.714(2
U+V
)+0.344(2
U
) (c.2)
R+B−G=Y+
0.714(2
U+
2
V
)+0.344(2
U+
2
V
) (c.3)
B+G=
2
Y+
0.714(2
U−V
) (c.4)
R+G=
2
Y+
0.344(2
V−U
) (c.5)
Equations (c.1) to (c.5) list a set of possible linear combinations of Equations (b.1) to (b.3). Note that Equations (c.1) to (c.5) use only two coefficients, namely 0.714 and 0.344, for matrix multiplications. Consequently, no more than two look-up tables may be used to convert luminance-chrominance color space signals to RGB color combination signals. Therefore, conversion from the luminance-chrominance color space to the RGB color space can be implemented using fewer than four look-up tables by converting the luminance-chrominance color space signals to the RGB color combination signals expressed as a function of predetermined linear combinations of the chrominance color space signals as defined by the appropriate conversion formulas, and by linearly combining the resulting RGB color combination signals to obtain the RGB color space signals.
According to one aspect of the invention, a method for converting digitized luminance-chrominance color space signals to digitized RGB color space signals comprises the steps of:
generating a plurality of predetermined linear combinations of the chrominance color space signals and at least one predetermined binary combination of the luminance color space signal;
providing a multiplexed multiplication unit which receives the linear combinations of the chrominance color space signals, the multiplexed multiplication unit including no more than two look-up tables which contain digitized transformation values for performing matrix multiplications of the linear combinations of the chrominance color space signals;
linearly combining the digitized transformation values outputted by the multiplexed multiplication unit and the binary combination of the luminance color space signal in a first predetermined manner to generate three RGB color combination signals; and
linearly combining the RGB color combination signals in a second predetermined manner to obtain the RGB color space signals.
According to another aspect of the invention, an apparatus for converting digitized luminance-chrominance color space signals to digitized RGB color space signals comprises:
a first combining unit for generating a plurality of predetermined linear combinations of the chrominance color space signals and at least one predetermined binary combination of the luminance color space signal;
a multiplexed multiplication unit connected to the first combining unit to receive the linear combinations of the chrominance color space signals therefrom, the multiplexed multiplication unit including no more than two look-up tables which contain digitized transformation values for performing matrix multiplications of the linear combinations of the chrominance color space signals;
a second combining unit connected to the multiplexed multiplication unit and the first combining unit, the second combining unit linearly combining the digitized transformation values outputted by the multiplexed multiplication unit and the binary combination of the luminance color space signal in a first predetermined manner to generate three RGB color combination signals; and
a third combining unit connected to the second combing unit, the third combining unit linearly combining the RGB color combination signals in a second predetermined manner to obtain the RGB color space signals.
In one embodiment, the multiplexed multiplication unit comprises first and second multiplexed look-up tables.
The first multiplexed look-up table includes: a first multiplexer having a plurality of data inputs which receive selected ones of the linear combinations of the chrominance color space signals, and a data output, the first multiplexer selecting each of the data inputs thereof in sequential first timing phases and providing data present at the selected one of the data inputs to the data output thereof; a first one of the look-up tables having an input connected to the data output of the first multiplexer, the first one of the look-up tables outputting the digitized transformation value corresponding to product of the data at the input thereof and a predetermined first coefficient; and a plurality of first data latches, each of which is connected to the first one of the look-up tables so as to latch outputs of the first one of the look-up tables during the sequential first timing phases, respectively.
The second multiplexed look-up table includes: a second multiplexer having a plurality of data inputs which receive selected ones of the linear combinations of the chrominance color space signals, and
Christie Parker & Hale LLP
Lee Michael
Tran Trang U.
LandOfFree
Method for performing plural matrix multiplication... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method for performing plural matrix multiplication..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method for performing plural matrix multiplication... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2449572