Image analysis – Image enhancement or restoration – Edge or contour enhancement
Reexamination Certificate
1998-08-17
2001-02-20
Lee, Thomas D. (Department: 2724)
Image analysis
Image enhancement or restoration
Edge or contour enhancement
C382S275000, C358S463000
Reexamination Certificate
active
06192162
ABSTRACT:
FIELD OF THE INVENTION
The invention relates to edge enhancing colored digital images.
BACKGROUND OF THE INVENTION
Currently, colored digital images are derived through various means including digital cameras and the digital scanning of film images. Many times the sharpness of a colored digital image is degraded by optical elements or by irregularities in the image sensor. For these reasons, it is often desirable to sharpen the color image after it has been converted to a digital representation. However, these digital images also tend to be noisy. For example, one noise source is sensor noise in both digital cameras and film scanners.
The problem with sharpening a noisy image is well known. Noise manifests itself as erroneous image fluctuations which are usually difficult to distinguish from the subtle image features which need sharpening. Thus, conventional sharpening processes amplify noise as well as image content. What is needed is a process that can sharpen image detail without further accentuating any noise present in the image.
Another consideration is that interpolated color images from digital cameras may have differences in edge sharpness depending on edge orientation and the details of the color interpolation algorithm. As a result, different regions of a colored digital image may require different levels of sharpening depending on edge orientation.
SUMMARY OF THE INVENTION
It is an object of the present invention to provide a method which enhances edges in digital color images.
This object is achieved by a method for edge enhancing a digital image having pixels comprising the steps of:
a) computing parameters for at least two orientations for each pixel in the stored digital image;
b) using the computed parameters to form a classifier value for each such orientation and using such classifiers to determine the pixel classification based on such classifiers; and
c) deriving a boost value according to the determined pixel classification.
ADVANTAGES
It has been previously stated that colored digital images are often noisy and in need of sharpening. Furthermore, different regions of a colored digital image from a digital camera need different levels of sharpening depending on the edge orientation. The present invention addresses these needs by classifying the edge content at each pixel as a function of the luminance channel of the image, and by applying the corresponding edge boost kernel.
Because each classification indicates a specific edge orientation, it is possible to create a corresponding edge boost kernel which is sensitive to image variations in one direction while simultaneously being insensitive to image variations in another direction. Thus, it is possible to sharpen edges while simultaneously diminishing the influence of noise.
Also, because classification identifies edge orientation, digital camera images can be given greater or lesser amounts of sharpening depending on properties of the color interpolation algorithm. The spatial frequency response of the edge sharpening process may also be adjusted as a function of edge orientation.
REFERENCES:
patent: 3971065 (1976-07-01), Bayer
patent: 5212740 (1993-05-01), Paek et al.
patent: 5652621 (1997-07-01), Adams, Jr. et al.
patent: 5719967 (1998-02-01), Sekine
Adams, Jr. James E.
Hamilton, Jr. John F.
Eastman Kodak Company
Lee Thomas D.
Owens Raymond L.
LandOfFree
Edge enhancing colored digital images does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Edge enhancing colored digital images, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Edge enhancing colored digital images will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2569080