Image analysis – Image enhancement or restoration – Intensity – brightness – contrast – or shading correction
Reexamination Certificate
1999-12-15
2004-11-02
Werner, Brian (Department: 2621)
Image analysis
Image enhancement or restoration
Intensity, brightness, contrast, or shading correction
C382S167000, C358S518000
Reexamination Certificate
active
06813389
ABSTRACT:
FIELD OF THE INVENTION
The invention relates generally to the field of image processing and more particularly to image processing systems which employ both a noise reduction processing step and the application of a tone scale function processing step.
BACKGROUND OF THE INVENTION
A typical digital imaging system involves three main components: a mechanism for generating the source digital imagery, a mechanism for processing the digital image data, and a mechanism for visualizing the imagery. As such, many digital imaging systems employ more than one image processing method, or algorithm, designed to enhance the visual quality of the final rendered output. In particular, two image processing methods of interest are methods that reduce the amount of noise present and methods for adjusting the tone scale of the processed images. In general, when these two types of image processing methods are employed in the same digital imaging system, each method is optimized separately. For some digital imaging systems, the application of a noise reduction method can affect the optimization of the tone scale adjustment method.
An example of noise reduction filter used in digital imaging systems is the Sigma filter, described by Jong-Sen Lee in the journal article Digital Image Smoothing and the Sigma filter, Computer Vision, Graphics, and Image Processing Vol 24, p. 255-269, 1983. This is a noise reduction filter that uses a non-linear pixel averaging technique sampled from a rectangular window about the center pixel. Pixels in the local neighborhood are either included or excluded from the numerical average on the basis of the difference between the pixel and the center pixel. Mathematically, the Sigma filter can be represented as
q
mn
=&Sgr;
ij
a
ij
p
ij
/&Sgr;
ij
a
ij
and
a
ij
=1 if |
p
ij
−p
mn
|<=&egr;
a
ij
=0 if |
p
ij
−p
mn
|>&egr;
where P
ij
represents the pixels in the local surround about the center pixel p
mn
, q
mn
represents the noise cleaned pixel, and &egr;represents a numerical constant usually set to two times the expected noise standard deviation. The local pixels are sampled from a rectangular region centered about the pixel of interest.
The Sigma filter was designed for digital image processing applications for which the dominant noise source is Gaussian additive noise. Signal dependent noise sources can easily be incorporated by making the &egr; parameter a function of the signal strength. However, for both signal independent and signal dependent noise cases the expected noise standard deviation must be known to obtain optimal results. Varying the &egr; parameter and the window size of the filter changes the strength of the Sigma filter. The amount of noise present in the processed image varies depending on the amount of noise in the starting image, the settings of the control parameters, and the structure of the image content. In low spatial activity regions, big changes to the noise levels are experienced while in highly structured regions very little change to the noise characteristics are experienced. The Sigma filter as a noise reduction filter can be used with a tone scale adjustment algorithm in the same system. However, the description given by Jong-Sen Lee makes no mention of using the Sigma filter in conjunction with other image processing algorithms in a digital imaging system.
In U.S. Pat. No. 5,134,573, Goodwin teaches a method for adjusting the tone scale for digitally scanned photographic film systems. It is claimed that this method improves the overall image contrast through the application of a tone scale function designed to linearize the photographic response of conventional photographic film products. Goodwin discloses a mathematical formula for constructing a tone scale function which relies on several control parameters. The mathematical formula was designed to accommodate a generalized photographic film product. The control parameters must be set according to the film response characteristics for a given photographic film product to achieve optimal results. One of the control parameters discussed is sensitive to the level of noise present in the digital image. No method for calculating the noise sensitive control parameter is mentioned. Only a range of values for the noise sensitive control parameter is offered based on the photographic film type. In addition, no mention of possible interactions with other image processing methods employed in a digital imaging system is mentioned either. If a noise reduction method is employed in a digital imaging system with this tone scale adjustment method, the optimal values of the noise sensitive control parameter will change.
U.S. Pat. No. 5,633,511 to Lee et. al. describes a method of tone scale adjustment involving constructing a tone scale function for a scanned radiographic digital imaging system. The method described involves a step which estimates the magnitude of noise present in the starting digital image as a function of pixel code value. Two distinct methods for estimating the image noise are discussed: 1) an off-line method of photographing a gray scale of uniform patches at different exposures, and 2) an on-line method of sampling uniformly exposed areas of the starting image. The first noise estimation method characterizes the noise properties of the photographic film product and scanner combination. The second method estimates the noise directly from the image pixel data of the starting image. No mention of possible interactions with other image processing methods employed in a digital imaging system is mentioned. If a noise reduction method is employed in a digital imaging system with this tone scale adjustment method, the noise characteristics of the starting digital image will be altered. The in-line method of estimating the noise characteristic will still work, however, this method is more computationally intense due to the spatial filtering required.
Most digital imaging systems do not coordinate the optimization of multiple image processing methods. One approach for optimizing multiple image processing operations is described in U.S. Pat. No. 5,694,484, issued Dec. 2, 1997, to Cottrell et al. Cottrell et al. propose an image processing system that automatically optimizes the perceptual quality of images undergoing a series of selected image-processing operations. The system consists of a set of image-processing operations, an architecture, and an intelligent control. These elements take into consideration profiles of source characteristics from which the images are generated, profiles of output device characteristics, and the impact that image processing operations (individually or in concert) will have on perceived image quality. Control parameters for the individual image processing operations are modified by optimizing an image quality metric (a single numerical quality) based on mathematical formulas relating objective metrics (such as sharpness, grain, tone, and color) with perceived image quality.
In the method described by Cottrell et al, there is no direct relationship between the individual control parameters for a noise reduction processing operation and a tone scale processing operation. The values for the individual control parameters are varied over useful ranges until the image quality metric achieves an optimal value. This method involves significant computation resources to evaluate the multitude of parameter permutations.
In practical digital imaging systems the computational resources available are limited. Consequently, for a digital imaging system which employs a noise reduction filter as one of the image processing operations, it can be advantageous to varying the amount of noise reduction to conserve computational resources. Varying the amount of noise reduction can change the optimal control parameter settings for tone scale adjustment algorithms if both are employed in the same digital imaging system. If a direct and simple relationship can be determined between a noise reduction fil
Gindele Edward B.
Lopez-Estrada Alex
Close Thomas H.
Werner Brian
LandOfFree
Digital image processing method and system including noise... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Digital image processing method and system including noise..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Digital image processing method and system including noise... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3328912