Image analysis – Pattern recognition – Feature extraction
Reexamination Certificate
1998-01-30
2001-02-06
Tran, Phuoc (Department: 2721)
Image analysis
Pattern recognition
Feature extraction
C382S260000, C382S264000, C382S266000, C345S182000, C345S440000
Reexamination Certificate
active
06185334
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Technical Field of the Invention
The present invention relates to inverse dithering, and more particularly, to a method for detecting an image edge within a windowed portion of a dithered image.
2. Description of Related Art
Dithering has long been a method to maintain the perceptual quality of continuous tone images in low bit depth video displays. The illusion of true continuous tone images can be produced on display devices incapable of providing a required number of amplitude levels necessary for achieving a full, continuous tone image by dithering. By dithering the images they may be displayed on systems having low bit-depth frame buffers while maintaining a reasonable image quality. The technique is also referred to as half-toning when the result produces a binary (black and white) image. Dithering enables a respectable display at a lower amplitude resolution by distributing quantization noise over small areas.
One process for improving the quality of a dithered image is referred to as inverse dithering. Inverse dithering strives to restore the original continuous tone nature of a predithered image. In inverse dithering, low-pass filters are used to reverse the effects of dithering such as graininess of the picture. However, use of low-pass filtering tends to blur the object edges of the image. Therefore, the inverse dithering process must ultimately achieve two conflicting goals. First, the effects of the original dithering must be smoothed, and at the same time the object edges or the high-frequency content of the original image must be maintained without undesirable blurring.
Inverse dithering may be discussed in a larger context of an image-rendering system. As illustrated in
FIG. 1
, a simplified image-rendering system consists of a dithering system
10
that receives as input, pixel values of an image I(x,y). The originally received image may be a monochromatic or a color image. The original image I(x,y) will have an amplitude resolution of q bits per pixel and the pixel values will be within the range of
I(xy)&egr;{0,1,2, . . . 2
q
−1}. (1)
The image I(x,y) is reduced in amplitude resolution to m bits per pixel by the dithering process. The dithered image I
d
(x,y) will consist of pixels having values within the range of
I
d
(x,y)&egr;{0,1,2, . . . 2
m
−1}. (2)
The dithered image I
d
(x,y) is processed by an image buffer
15
which in many computer display systems would typically comprise a frame buffer. However, the image buffer
15
may comprise any number of other mechanisms for holding or transporting an image such as a communication channel. The image buffer
15
actually drives the necessity for amplitude reduction of the original image I(x,y) due to the limited amount of memory in the buffer for storage or the narrow bandwidth of a communication channel.
The inverse dithering system
20
receives the dithered image I
d
(x,y) and produces a reconstructed image I
r
(x,y) having a higher amplitude resolution of q″-bits per pixel. The amplitude resolution of the original and reconstructed images may be the same, but this is not required. The reconstructed image will have pixel values within the range
I
r
(x,y)&egr;{0,1,2, . . . , 2
q″
−1}. (3)
A variety of methods have been developed for performing inverse dithering. The majority of these methods have been focused on the specialized case of inverse half-toning or, in other words, recovery of full gray scale from binary images. In one technique, a gray scale image is constructed from a binary image by means of a statistically generated lookup table. This method utilizes statistics generated during the original dithering process to perform the. statistical analysis necessary to generate the lookup table. The statistical analysis requires complex calculations, and the storage of the lookup table greatly increases the memory requirements.
Another method utilizes cascading of adaptive run-length algorithms with statistical smoothing and impulse removal. A class of iterative, nonlinear algorithms have been proposed that are based on the theory of Projection Onto Convey Sets (POCS). This technique involves performing multiple iterations of an inverse dithering process until a final inverse dithered image is determined. This requires a great deal of processing due to the multiple iterations. Other methods have used information generated by a wavelet decomposition to perform the inverse dithering process.
There are also methods for reconstructing a dithered color image referred to as “color recovery”. The color recovery technique requires a renormalization process that increases the overhead requirements for processing the inverse dithering algorithm.
Some of the aforementioned techniques require prior knowledge of the dither mask used to dither the original image. Also, most of these methods involve highly complex calculations and require a great deal of memory in order to perform the inverse dithering processes. These requirements limit the ease and speed in which an inverse dithering process may be performed. Thus, there has arisen a need for a method of inverse dithering that is substantially less computationally complex and requires less memory resources than those of methods presently utilized.
OBJECTS OF THE INVENTION
Accordingly, it is an object of the present invention to provide a method and apparatus for inverse dithering image which has been dithered according to any one of a plurality of dithering methods.
It is also an object of the present invention to provide a technique for determining whether an image edge exists within a region of support of a filter processing the dithered image.
It is yet another object of the present invention to provide a technique for detecting an edge of a dithered image based upon predetermined data.
It is still further an object of the present invention to provide further advantages and features, which will become more apparent to those skilled in the art from the disclosure, including a preferred embodiment, which shall be described below.
SUMMARY OF THE INVENTION
The present invention is directed to the above need. A method in accordance with the present invention includes the steps of selecting a first filter from a set of filters, where the selected first filter has a first filter coefficient and is operable on a first region of support, and then determining that an image edge exists within the first region of support. In response to determining that an image edge exists within the first region of support, a second filter is selected from the set of filters, where the selected second filter has a second filter coefficient and is operable on a second region of support and the second region of support is smaller than the first region of support. Next, it is determined if an image edge exists within the second region of support, and in response to determining that no image edge exists within in the second region of support, the dithered image is reconstructed using the second filter coefficient.
An article of manufacture in accordance with the present invention includes a computer readable storage medium and a computer program stored on the storage medium, where the computer program is configured to be readable from the storage medium by a processor thereby causing the processor to operate (i) to select a first filter from a set of filters, where the selected first filter has a first filter coefficient and is operable on a first region of support, (ii) to determine that an image edge exists within the first region of support, (iii) to select a second filter from the set of filters in response to determining that an image edge exists within the first region of support, where the second selected filter has a second filter coefficient and is operable on a second region of support and where the second region is smaller that the first region, (iv) and to reconstruct the dithered image using the second filter coefficient, in response to determining that no i
Cheung Shiufun
Iyengar Giridharan
MacNamara Robert
Ulichney Robert A.
Compaq Computer Corporation
Leah Sherry Oppenheimer Wolff & Donnelly LLP
Mariam Daniel G.
Tran Phuoc
LandOfFree
Method for reconstructing a dithered image 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 reconstructing a dithered image, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method for reconstructing a dithered image will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2565339