Image analysis – Image enhancement or restoration – Variable threshold – gain – or slice level
Reexamination Certificate
1998-01-21
2001-01-30
Mehta, Bhavesh (Department: 2721)
Image analysis
Image enhancement or restoration
Variable threshold, gain, or slice level
C382S264000, C382S224000, C358S426010
Reexamination Certificate
active
06181829
ABSTRACT:
FIELD OF THE PRESENT INVENTION
The present invention relates generally to a system for processing document images, and more particularly, to an improved method of image processing the document images utilizing a fuzzy logic classification process.
BACKGROUND OF THE PRESENT INVENTION
In the reproduction of images from an original document or images from video image data, and more particularly, to the rendering of image data representing an original document that has been electronically scanned, one is faced with limited reflectance domain resolution capabilities because most output devices are binary or require compression to binary for storage efficiency. This is particularly evident when attempting to reproduce halftones, lines, and continuous tone (contone) images.
An image data processing system may be tailored so as to offset the limited reflectance domain resolution capabilities of the rendering apparatus, but this tailoring is difficult due to the divergent processing needs required by different types of images which may be encountered by the rendering device. In this respect, it should be understood that the image content of the original document may consist of multiple image types, including halftones of various frequencies, continuous tones (contones), line copy, error diffused images, etc. or a combination of any of the above, and some unknown degree of some or all of the above or additional image types.
In view of the situation, optimizing the image processing system for one image type in an effort to offset the limitations in the resolution and the depth capability of the rendering apparatus may not be possible, requiring a compromised choice which may not produce acceptable results. Thus, for example, where one optimizes the system for low frequency halftones, it is often at the expense of degraded rendering of high frequency halftones, or of line copy, and visa versa.
To address this particular situation, “prior art” devices have utilized automatic image segmentation to serve as a tool to identify different image types or imagery. For example, in one such system, image segmentation was addressed by applying a function to the video, the output of which was used to instruct the image processing system as to the type of image data present so that it could be processed appropriately. In particular, an auto-correlation function was applied to the stream of pixel data to detect the existence and estimate the frequency of halftone image data. Such a method automatically processes a stream of image pixels representing unknown combinations of high and low frequency halftones, contones, and/or lines. The auto-correlation function was applied to the stream of image pixels, and for the portions of the stream that contain high frequency halftone image data, the function produced a large number of closely spaced peaks in the resultant signal.
In another auto-segmentation process, an auto-correlation function is calculated for the stream of halftone image data at selected time delays which are predicted to be indicative of the image frequency characteristics, without prior thresholding. Valleys in the resulting auto-correlated function are detected to determine whether a high frequency halftone image is present.
An example of a “prior art” automatic segmentation circuit is illustrated in FIG.
6
. The basic system as shown in
FIG. 6
is made up of three modules. Input information stored in a data buffer
10
is simultaneously directed to an image property classifying section
20
, the first module, and an image processing section
30
, the second module. The image property classifying section
20
, is made up of any number of submodules, (e.g. auto-correlator
21
and discriminator
22
), which determine whether a block of image pixels stored in the data buffer
10
is one type of imagery or another, (e.g. halftone, line/text, or contone). In parallel with the image property classifying section
20
, the image processing section
30
is made up of any number of sub-processing sections, (e.g. high frequency halftone processor
31
low frequency halftone processor
32
, line/text processor
33
, or contone processor
34
), which perform image processing operations on the same block of image pixels as section
20
. Each image sub-processing section performs image processing operations that are adapted to improve the image quality of a distinct class of imagery. The third module, control section
41
, uses the information derived from the image classifying section
20
, to control the image processing section
30
. In other words, the control section
41
acts like a multiplexer and selects the proper processed image data according to the image classification determined by the image classifying section
20
.
The decision as to what class of imagery image data belongs to is typically binary in nature. For example, in a conventional image segmentation scheme image property classifying section
20
classifies image data as one of three classes of imagery, (high frequency halftone, low frequency halftone, or contone). Depending on those classification, image data is processed according to the properties of that class of imagery is selected, (either low pass filter and re-screening if it's a high frequency halftone, threshold with a random threshold if it is a low frequency halftone, etc.). Also, assuming that the decision as to which of the three classes of imagery image data belongs is based on a single image property, the peak count of the input image data, the resulting image classification decision of the peak count image property is made by thresholding the peak count into three classes of imagery.
Consequently, the control section
40
decides the type of image processing the image data requires depending on the decision made by the classification section
20
. Thus, the output of classification section
20
is quantized to one of three possibilities. The control section
40
selects the output from one of the three image sub-processing sections based upon this classification.
Based on the nature of conventional image classification systems, the classifying information, gathered over a context of many pixels, changes gradually. But in the process of comparing this classifying information with a classification threshold one could create abrupt change in the classes. This abrupt decision making, which produces a forced choice among several distinct alternative choices, is a primary reason for the formation of visible artifacts in the resulting output image. Most transition points or thresholds are selected so that an image can be classified as one class of imagery with a high degree of certainty; however, those classes of imagery that cannot be classified with such certainty have multiple transition points or a transition zone.
Using only one point to define a transition zone results in the formation of visible artifacts in the resulting output image if the output image spans in the transition zone. Although it is possible to shift or make the transition zone narrower so that there is less chance that an image falls into the zone, there exists limitations on how narrow the zone can be made. The narrowing of the transition zone is the decreasing of noise and/or variation in the information used to classify so as to narrow the area over which classification is not “certain”, resulting in less switching between classifications.
Moreover, the classification of real images covers a continuum from well below to well above thresholds between classifications. This means that there are areas of an image which are, for example, just above a threshold. Variations in the gathered (lowpass filtered) information due to “flaws” in the input video or ripple due to interactions between the area of image being used for the classification process and periodic structures in the input video results in areas falling below the threshold. With discrete classification, this results in a drastically different classification, thereby resulting in artifacts in the rendered image.
Thus, it is desirable
Clark Raymond J.
Schweid Stuart A.
Shiau Jeng-Nan
Williams Leon C.
Mehta Bhavesh
Nickerson Michael J.
Patel Kanji
Xerox Corporation
LandOfFree
Method and system for classifying and processing of pixels... 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 and system for classifying and processing of pixels..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method and system for classifying and processing of pixels... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2503822