Facsimile and static presentation processing – Static presentation processing – Attribute control
Reexamination Certificate
2000-04-04
2004-04-27
Rogers, Scott (Department: 2626)
Facsimile and static presentation processing
Static presentation processing
Attribute control
C358S003260, C382S260000
Reexamination Certificate
active
06728003
ABSTRACT:
FIELD OF THE INVENTION
The invention relates generally to the field of digital image processing and, more particularly, to a method of compensating for MTF degradations in the imaging chain.
BACKGROUND OF THE INVENTION
Physical devices, such as capture or output devices in the imaging chain degrade the frequency content of the original image signal. Examples of capture devices include a CCD (charge coupled device), and a scanner for scanning a film negative. Examples of output devices are printers and displays in an imaging chain. For a linear shift-invariant (LSI) system, the specific nature of the degradation imposed by the device is referred to as the Modulation Transfer Function (MTF) of the device. The MTF basically specifies the frequency response of the device. Strictly speaking, the MTF specifies the relative attenuation or boost of the modulation of sinusoidal inputs to a given system or component. (MTF does not give phase information.) Each device or component of an imaging chain has an associated MTF. In addition, the MTF of several components may be cascaded (multiplicatively) by techniques well known in the art (see for example commonly-assigned U.S. Pat. No. 5,696,850.) Thus, the MTF of multiple devices or an entire system may be determined by cascading the MTF of the appropriate devices.
It is desirable to compensate for the MTF losses caused by one or more devices in the imaging chain at the most appropriate point. The most appropriate place to compensate for the MTF of the device is at the point that the device is used. For instance, if a CCD is used to capture an image in a digital camera, the MTF of the CCD should be compensated just after the image is captured. Changes to the original resolution and the inclusion of devices with additional MTF losses to the imaging chain make the compensation for the MTF of a device a much more difficult problem if the device compensation is separated from the location of the device in the image chain.
Currently, many image processing systems sharpen in only one place (often at the printer.) This one stage of sharpening attempts to compensate for the MTF losses of every device contained in the image chain. Additionally, part of the single stage sharpening is designed to boost the image detail for human preference. In this type of system, the change of any single component of the image chain may require the re-optimization of the sharpening. However, if the MTF of each device is compensated for, then the change of a single component of the image chain merely requires the determination of the MTF compensation for the new component by a pre-defined process.
In designing an MTF compensation scheme, the term “compensation” must be defined. In the past, a restrictive definition of compensation has been used and it requires that the sequential combination of the device and the device compensation has an MTF of 1.0 for all frequencies. This strict device compensation would be accomplished with an inverse filter, such as described by Gonzalas and Woods, in
Digital Image Processing
, Addison-Wesley Publishing Company, 1992, pp. 270-272. For example, the inverse filter is the design goal specified by commonly-assigned U.S. Pat. No. 5,696,850 when describing a method of performing a single sharpening operation to compensate for both input and output device. In practice, the application of an inverse filter would result in many problems, especially because the gain of the inverse filter would approach infinity where the MTF of the device approaches 0.0. Other factors making it impossible to attain good results from the inverse filter include noise, measurement error, limited device dynamic range, and device nonlinearities.
Because of the difficulties associated with strict device compensation, a more practical approach must be taken. Non-strict device compensation is a process by which the MTF of a device is restored within a reasonable delta of 1.0 for as great a frequency range as possible, giving preference to the lower frequencies. Note that the definitions of the terms “reasonable delta” and “frequency range” may vary.
Several non-strict device compensation methods can be found in the literature. Ohkouchi and Suzuki in U.S. Pat. No. 4,517,607 and Kamiya in U.S. Pat. No. 4,817,181 speak of MTF compensation, but the digital filter is held constant for any device. Thus, their MTF compensation does not at all depend upon the MTF of the device. Additionally, Takashi et al. in U.S. Pat. No. 5,144,686 also describe an MTF compensation that is independent of device MTF. While methods of sharpening images independent of the device MTF such as mentioned in this paragraph are certainly useful for producing higher quality images, it would be very useful to have a deterministic method of arriving at the non-strict device compensation based upon knowledge of the device MTF.
Sumi, U.S. Pat. No. 5,191,439 designed a sharpening system for compensating for device MTF. Although the combination of the printer and the compensation is considered, Sumi fails to describe a deterministic method of arriving at an acceptable non-strict device compensation based upon the device MTF.
Ishii et al., in U.S. Pat. No. 5,978,522 describes a method of modifying the sharpness characteristic of a digital image similar to the inverse filter described by commonly-assigned U.S. Pat. No. 5,696,850. However, Ishii utilizes an aim specific to the source of the image, or a user's preference, rather than an aim magnitude frequency response of 1.0 for all frequencies. This method has no guarantees that the compensation will not require high gains which will greatly amplify the noise in the system. In addition, this method may generate a compensation filter that is very large, since the filter is generated directly from an inverse Fourier transform of a filter aim. Also, each frequency is considered of equal importance to the correction.
The Wiener filter (commonly known in the art, for example described by Gonzalas and Woods, in
Digital Image Processing
, Addison-Wesley Publishing Company, 1992, pp. 279-282) is actually an attempt to apply a non-strict device compensation. The Wiener filter requires knowledge of both the magnitude and phase of the frequency response of the device, as well as the power spectrum of both the signal and the noise (although approximations can be made if the power spectra are unknown.) The Wiener filter uses the signal-to-noise-ratio (SNR) to limit the gain applied to any specific frequency. Generally, in a typical imaging device, the SNR of the higher frequencies is lower than the SNR of the low frequencies. As a result of the low SNR, the magnitude of the frequency response of the high frequencies is not restored to 1.0. Additionally, the low frequencies are generally allowed to be more accurately restored, assuming that the SNR is not low. Unfortunately, the Wiener filter is computationally intensive, requiring the use of a Fourier transform of the image (although the correction spectrum need be computed only once per device.) However, the desirable traits of the Wiener filter should be preserved when designing a strategy for non-strict device compensation.
SUMMARY OF THE INVENTION
It is an object of the present invention to provide an improved method for compensating for the MTF degradations of one or more devices in the imaging chain.
It is a further object of the invention to limit the overall gain of the compensation so that noise is not excessively amplified.
These objects are achieved in a method for processing a digital image channel which is part of a digital image to compensate for MTF of one or more devices in an imaging chain from capture to printing comprising the steps of:
(a) providing the MTF of the one or more devices in the imaging chain;
(b) providing a gain factor, using the MTF to provide an aim response;
(c) providing a filter from the aim response; and
(d) applying the filter to the digital image channel to provide a new digital image channel.
It is an advantage of the present invention that the MTF in one or more de
Gallagher Andrew
Parada Robert J.
Owens Raymond L.
Rogers Scott
LandOfFree
Method of compensating for MTF in a digital image channel 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 of compensating for MTF in a digital image channel, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method of compensating for MTF in a digital image channel will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3224895