Computer graphics processing and selective visual display system – Computer graphics processing – Graphic manipulation
Reexamination Certificate
2000-03-24
2003-01-14
Brier, Jeffery (Department: 2672)
Computer graphics processing and selective visual display system
Computer graphics processing
Graphic manipulation
C341S140000, C710S033000, C375S240010, C348S254000
Reexamination Certificate
active
06507347
ABSTRACT:
FIELD OF THE INVENTION
This invention relates to selected data compression for digital pictorial information and, in particular, the supply of digital information to digital display screens. This includes a variety of forms of digital display screen including large screen LED displays, LCD displays, LCD projectors, plasma television and similar forms of apparatus.
BACKGROUND TO THE INVENTION
Original video and television technology utilized analogue signals for the recording, transmission and driving of displays.
The cathode ray tubes (CRTs) still predominate the worldwide market for display apparatus in a form of televisions or similar. Therefore, it will be sometime before digital displays predominant to the extent that original data captured in the form of video, film or other analogue recording mechanisms are entirely replaced by digital mechanisms facilitating direct digital recording to digital transmission and reproduction.
The operation of a CRT is such that the brightness or intensity of a particular pixel on a screen is determined by the quantity of electrons accelerated by the cathode ray tube onto that point of the screen. However, the power or voltage supplied to the cathode ray tube leads to a non-linear response in the number of electrons accelerated.
This lack of linearity is well known and referred to as the gamma effect.
Different transmission or reproduction standards set different gamma functions to account for this lack of linearity. The gamma function not only varies between industry standards such as NTSC or PAL but also with proprietary brands of monitors and similar display technologies.
Current recording technologies can involve analogue cameras which themselves work on a non-linear basis and although they may need to be compensated to the particular gamma function of the display on which the signal is reproduced, there is little identifiable loss of definition during this compensation process.
If information is recorded in a digital form, various intensities are recorded by discreet binary numbers. For example, if the information is recorded using 8-bit technology, different levels of intensity are graduated according to the 256 possible binary numbers available in an 8-bit binary number. This is substantially a linear representation of the intensity. Therefore, to allow such digital information to be displayed utilizing conventional non-linear technology, it is necessary to add a gamma function to this information to provide a non-linear correlation between the intensity and the individual steps from each bit of information.
Regardless of the method of recording the data, as presented for reproduction, the data is likely to come in analogue or digital form having a gamma function. It is only if the recording process has been performed directly on digital equipment and no gamma function has been added as it is directly intended for use on a digital display that such a function may not exist. Transmissions by broadcast also benefit from a lack of linearity in reducing noise.
The result of this is that most digital displays need to work with such non-linear data.
When an analogue signal is received to represent pictorial information for a digital display, one of the first steps is to convert the analogue signal into digital information for processing. Whether converted from analogue or initially digital, the signal then needs to be linearized to remove the gamma function and provide the data in a manner that the digital display, that operates substantially linearly, will correctly represent the intended image. If the processed signal is supplied directly to the digital display, apparent visual distortions can occur in the low intensity colours. For low intensity colours, a single step in the binary information, once the gamma function has been removed, can lead to a distinct visual change in intensity. This leads to an effect referred to as mach banding where regions of low intensity colour approaching black can show distinct bands of colour where a single binary step in the digital information needs to represent a large percentage change in colour from the original non-linear signal. The reverse is true at high intensity colours so that in any high signal intensity colour that has the gamma function removed can easily be represented on a digital display. Large percentage errors between the anti-gamma digital representation and the non-linear signal occur at low intensity colours with minimal errors at the higher intensity colours.
To overcome this effect, different approaches have been taken in the past.
One approach is to use some error diffusion in the representation of the digital display. In effect, in an area where a band boundary would normally be apparent, some pixels in the lower intensity band are provided with the value of the higher intensity band and vice versa to provide the appearance of graduation in the intensities rather than any distinct step.
Although this effect can work for some digital displays, other digital displays utilize significantly larger and more apparent pixels. This may particularly be the case for large indoor or outdoor LED display screens where each individual pixel is a substantially larger unit and can be visually apparent on its own. Also, the total number of pixels in the display may be less in some displays making each individual pixel more important to the overall image. The use of error diffusion on such displays leads to a loss of definition rather than simply overcoming the banding.
The alternative course of action that is taken with most LED displays and indeed with other digital displays is to increase the accuracy of the digital information subsequent to removal of the gamma function by using an increased number of bits.
Typically, if an 8-bit digital signal is used as a direct representation of the nonlinear signal, an anti-gamma function needs to be applied to linearize the non-linear data. The application of this function loses some accuracy as the smaller binary numbers lose definition. For example, if an 8-bit signal provides a representation of the numeral
16
representing a small value of the 256 possible graduations, the application of the anti-gamma function may indicate that the true value should be, perhaps, 0.65. Such a number cannot be represented by a subsequent 8-bit binary number. Hence the output may simply be “1”. Due to the nature of the anti-gamma calculation, the number
25
may normally equal 1.45 but still need to be represented as “1” in the recalculated linearized 8-bit data. This occurs only on those lower intensity values. Generally, the data will be truncated or rounded to lose significant digits.
Greater accuracy can be employed by outputting the 8-bit data, once the anti-gamma function has been applied as a 10-bit binary number to allow a greater degree of accuracy on these low intensity colours.
Of course, some systems utilize higher bits from the outset although such apparatus is naturally more expensive as greater processing abilities are needed throughout the entire system. The use of 8-bit technology has become a standard for the more economic forms of digital pictorial information such as a standard video.
The loss of definition in colours once the anti-gamma function is applied means that the original 8-bit information comprising some 256 discreet levels is reduced upon output. For an anti-gamma function of 2.2 which is substantially the standard for NTSC signals, only approximately 184 discreet output colours are possible using an 8-bit output. However, if the 8-bit input receives the anti-gamma function and is provided as a 10-bit output, approximately 233 colours are possible. With higher bit outputs, even more colours are available with a 16-bit output providing substantially the same 256 discreet colours which are considered sufficient to substantially correspond to an original analogue signal.
The difficulty in providing an increase in data bits once the anti-gamma function has been applied is that all the downstream equipment must similarly be able to transmi
Brier Jeffery
Jacobson & Holman PLLC
Lighthouse Technologies Ltd.
Yang Ryan
LandOfFree
Selected data compression for digital pictorial information does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Selected data compression for digital pictorial information, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Selected data compression for digital pictorial information will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3039037