Television – Camera – system and detail
Reexamination Certificate
1999-10-29
2004-04-13
Christensen, Andrew (Department: 2615)
Television
Camera, system and detail
C348S342000, C348S254000
Reexamination Certificate
active
06720992
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Technical Field
The invention relates to a method which allows considerable improvement in the recording and reproduction of the light/dark contrast in digital images. The invention is thus of major interest for digital image processing (imaging).
2. Description of the Prior Art
The typical image processing process will be explained, first of all, using a flowchart to provide better understanding of what is stated in the following text. In this case, digital imaging is broken down into three blocks (FIG.
1
). As is evident without any difficulty from the inscriptions on the figure, the scanning of the original to be imaged is carried out in Block
1
by means of an image sensor or scanner. The image data obtained in this way are then processed in a next step (Block
2
). Two different methods may be used in this case:
1.) Interactive image processing, in which the image can be corrected or edited as required by manual processing of the data record on a screen, and/or
2.) Automatic image processing, in which standardized computation runs are possible, which in turn allows very short process times. In this way, image characteristics such as color neutrality, brightness, and image contrast can be optimized.
Finally, the image data records which have been improved by one method or the other are stored and are then available for the image to be output via various imaging systems, which process the images in the form of pixels (Block
3
). These include, for example, the CRT screen, exposure of photographic materials by pixels, cinematic projection of digitally recorded “movies” by means of laser projectors or mirror chips (DLP), newspaper printing or printing out on inkjet printers, to name but a few.
While the scaling technique for recording and for pixel image reproduction must be matched to the respective application purposes, it is possible to use largely standard techniques for the image processing activities that occur in between. For example, time consuming interactive image processing is predominantly used for individual images of All types (for example in the advertizing field), while image processing is becoming increasingly important, for example for large-scale photographic copying, in the TV or amateur-video area or for the processing of digitized films. In this case, automatic contrast adaptation, in particular, plays an important role in addition to adaptation of the image brightness and color reproduction.
Contrast adaptation is therefore required, in particular, because natural image reproduction is adversely affected, above all, by two contrast problems. On the one hand, most reproduction media, for example photographic paper, printed paper, or even screens, cannot always satisfactorily reproduce relatively major original contrasts and, on the other hand, it is often necessary to work with the respectively currently available lighting conditions (available light) during recording, that is to say lighting conditions which are unbalanced and, in some circumstances, change rapidly. With regard to image reproduction, both problems can lead to scenes with, for example, overexposed sky sections and/or black shadows with no image information in them.
For some years, methods have been known which are suitable for reducing such image defects by subsequent image data processing. For example, by linking the image data to data records which contain only low local frequencies from the original (i.e. blurred masking), it is possible to reproduce excessively bright image parts attenuated and shadowed areas brightened in the processed image data records of the original. Thus, as a fundamental rule for improving image contrast the contrast of the relatively large—area image sections (i.e. the low local frequencies) reduced, while the detail contrast (important for image brilliance) remains unaffected.
By superimposing a number of local frequency extracts, each having different contrast in one mask, it is even possible to roughly copy the visual acuity of the eye. This blurred mask type, which is characterized by the name multilayer mask, can be used in a corrective sense in the same way for reproduction of originals. Applied to the image data record supplied by the image sensor—it allows images to be produced with a contrast impression that is still more natural.
Blurred masking with the aid of LCDs for mass copying of photographic originals has already been disclosed in OS DE 4040498 [sic]—likewise with the aim of improving the handling of exposure problems in particularly high-contrast originals, methods have also been described for the photographic area which allow the sensitivity of flat photographic material to be attenuated or emphasized in an analogous manner to rough brightness distribution of the original to be imaged. Using such a method; the details in the shadow areas and in the highlights could then also be reproduced more clearly, that is to say with sufficiently good exposure (See DE 196 32 429).
The use of blurred masks (including the multilayer mask mentioned above) both in the video sector and in conjunction with influencing the sensitivity of image sensors in areas, in an analogous manner to masks, have been disclosed in Patent Specification DE 19713 648.
The digital contrast adaptation described in the above mentioned documents generally works in such a way that an attempt is made to attenuate or to amplify the image signals (taken from the image sensor) the light and dark image areas retrospectively in areas. It is obvious that such image processing methods can lead to really good results only if the highlighted and shadowed areas which may be present in the data records are correctly exposed and have good information content. As has already been addressed in the last-mentioned Patent Specification (DE 19713 648), this would best be ensured by using image sensors whose light sensitivity can be adapted in areas to brightness conditions in the highlighted and shadowed zones.
Such image sensors now have acceptable resolution and have now been marketed for some time under the name CMOS active pixel sensors. The sensitivity (or the exposure time) of the pixels in these image sensors can be programmed individually. Their use promises the capability to achieve a previously unknown quality by using automatic processes to form image data records. At the same time considerable proportion of the “work” of the image processing station (
FIG. 1
, Block
2
) would be carried out by the active pixel digital camera itself.
SUMMARY OF THE INVENTION
The aim of the present application is to describe suitable logic and actuation principles for such image sensors (referred to for short as AP image sensors in the following text), by means of which it is possible to obtain image data records with balanced contrast.
DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT OF THE INVENTION
In the case of the first method (FIG.
2
), also called the 1-chip solution for short, the original (
1
) is first of all imaged in focus on the AP image sensor (
2
). In this case, the distribution of its pixel sensitivity is intended to be roughly uniform. Digitized image data from this image are then supplied according to Step
1
to a computer (
3
), in which a blurred monochromatic positive of the original is calculated by means of blurring algorithms.
In the simplest case, this blurred image data record of the original (,“low-pass filter”) is a low-frequency, local frequency area extract from the original. Its corrective effect depends not only on the amount of blurring but also on the contrast given to it in the calculation, that is to say the density difference between its brightest and darkest area. For simplicity, this data record is also called a “positive blurred mask.”
REFERENCES:
patent: 5051902 (1991-09-01), Hishinuma
patent: 5319719 (1994-06-01), Nakazawa et al.
patent: 5363209 (1994-11-01), Eschbach et al.
patent: 5774599 (1998-06-01), Muka et al.
patent: 6069979 (2000-05-01), VanMetter
patent: 6094511 (2000-07-01), Metcalfe et al.
patent: 614681
von Stein Werner Ritter
Zahn Wolfgang
Glenn Michael A.
Glenn Patent Group
Harris Tia M.
LandOfFree
Method for improving the contrast reproduction of digitized... 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 improving the contrast reproduction of digitized..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method for improving the contrast reproduction of digitized... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3245882