Facsimile and static presentation processing – Natural color facsimile – Color correction
Reexamination Certificate
1999-05-10
2002-08-06
Coles, Edward (Department: 2722)
Facsimile and static presentation processing
Natural color facsimile
Color correction
C348S234000, C348S265000, C345S156000, C358S482000, C358S483000
Reexamination Certificate
active
06429953
ABSTRACT:
FIELD OF THE INVENTION
This invention relates to document scanning for copier, scanner, and facsimile, etc., and specifically to a method and system for increasing resolution therein without the requirement of providing more sensing elements.
BACKGROUND OF THE INVENTION
Prior art scanners generally employ a red-green-blue (RGB) sensor array to capture an image as shown in
FIG. 1
, generally at
10
. Each array has three channels, including a red channel
12
, a green channel
14
and a blue channel
16
. All three channels have the same number of detector, or sensor, elements, such as sensors
18
,
20
and
22
. The three channels each generate a signal, which signals are combined to form a color image. This approach does not take advantage of the human vision system (HVS), wherein the bandwidth for chrominance signals is significantly lower than that of the luminance signal. A full resolution RGB scanner captures redundant information. In fact, the first step of many image processing algorithms is to remove this redundancy by converting the RGB signal to luminance-chrominance-chrominance (LCC), and sub-sampling the chrominance channels reduce this redundancy, as by JPEG compression, or other well known techniques as used in color television or copiers.
Most commercially available scanners capture color images using three sensors: red (R), green (G), and blue (B). For most desktop scanners, this is accomplished with a 3-row CCD sensor, wherein each sensor row has a number of sensors therein. Each row is offset in the slow scan direction, or sub-direction, i.e., the direction of scan bar movement, by a few scanlines, and each row of sensors is responsible for capturing only one color. The captured data are recombined digitally to form an RGB image. The three RGB sensors typically have the same number of sensors which are aligned in the slow scan direction, indicated by arrow
24
in FIG.
1
. Scanning resolution is determined by the spacing between two adjacent sensor elements and the optical magnification. In order to achieve high resolution, more sensor elements are used, which results in an increased resolution from the original 75 dpi to 300 dpi and now, in some instances, up to 600 dpi, however, adding more sensor elements is expensive, and other resolution enhancement methods have been explored.
A number of patent references are known which use color multiplexing to improve resolution. Nearly all of those are for 2D video or digital still cameras where color filter arrays (CFA) are used to derive 2D color images from a single 2D CCD sensor. However, because of the inherent difficulty associated with sampling in a 2D arrangement, it is difficult, if not impossible, to implement super sampling of luminance in this type of device.
U.S. Pat. No. 4,575,769 to Arnoldi, for Variable resolution scanning in line scan optical imaging system, granted Mar. 11, 1986, discloses a method of improving resolution by advancing a document past a horizontal image scanner in defined sequences of half and full steps, half steps causing the document to be feed wise advanced by one four-hundredth of an inch, as in 400 dpi mode, full steps causing the document to be feed wise advanced by one two-hundredth of an inch (as in 200 dpi mode). This technique improves resolution only in the slow scan direction, not the fast scan direction, i.e., parallel to the CCD array.
U.S. Pat. No. 4,877, 310 to Seachman et al., for Electronically variable MTF filter for image sensor arrays, granted Oct. 31, 1989, describes placement of a MTF filter between a lens and an imaging array to reduce image modulation to blur the image prior to further processing.
U.S. Pat. No. 5,045,932 to Sharman, et. al., for Method and apparatus for generating a high definition electronic signal from a line scan of a color original, granted Sep. 3, 1991, describes a motion picture film scanner that generates a high definition television signal from the combination of a high definition luminance component and a plurality of lower definition color components. The color sensors can have large sensor apertures resulting in better signal-to-noise ratio. This system uses four sensor arrays, and is considered to be too complex and too expensive for low cost desktop scanners.
U.S. Pat. No. 5,159,469 to Takagi, for Moire removing device for image scanner, granted Oct. 27, 1992, describes a movable filtering mechanism to focus and defocus an image at an array plane.
U.S. Pat. No. 5,262,631 to Yamamoto, et al., for Color image reading apparatus, granted Nov. 16, 1993, employs a two-row sensor, one to capture green and the other to capture red and blue with alternative red and blue sensors. The green row is shifted one-half pixel in relation to the red/blue row. This device is a RGB capturing device, and does not have the bandwidth advantage associated with LCC capture.
U.S. Pat. No. 5,767,987 to Wolff, et. al., for Method and Apparatus for combining multiple image scans for enhanced resolution, granted Jun. 16, 1998, discloses a method of generating an enhanced resolution image from multiple low resolution images. One of the images is used as a prototype image, which is interpolated to higher resolution. The other images are then registered with the prototype image and the pixel values are changed iteratively based on the images data of the additional images. This technique works because the inherent jitter error of a scanner introduces more sampling points, which may be used to improve resolution. This method requires multiple scans and intensive processing.
U.S. Pat. No. 5,773,814 to Phillips, et al., for Sensor assembly providing gray scale and color for an optical image scanner, granted Jun. 30, 1998, uses Y and W interleaved on one ID array and G and C on another. The stated purpose of this technique is to optimize light collection and maximize SNR to allow for faster scan times, not to improve the image resolution. There is no shift, or misalignment, between sensor arrays.
All these prior art approaches address the resolution problem differently. Most of them do not take advantage of the human color vision, where chrominance bandwidth is significantly less than that of luminance, or used the properties of the HVS as a way of reducing bandwidth, instead of enhancing resolution.
SUMMARY OF THE INVENTION
A method for super-resolution scanning in a image capturing device includes capturing a first color component of the image at one-half of a full resolution, capturing a first luminance component at one-half of the full resolution, and generating a first color-luminance output signal therefrom; capturing a second luminance component of the image at the full resolution and generating a luminance output signal therefrom; capturing a second color component of the image at one-half the full resolution, capturing a third luminance component of the image at one-half the full resolution, and generating a second color-luminance output signal therefrom; converting the first and second luminance components of the first and second color-luminance luminance output signals and the third luminance component signal to a luminance component signal of 2X full resolution; and processing the full resolution output signals and the 2X full resolution output signal to form a combined output signal.
A system for super-resolution scanning in a image capturing device includes a sensor array having a first row with first color pixel capturing elements which alternate with first luminance pixel capturing elements; a second row having second luminance pixel capturing elements; a third row with second color pixel capturing elements which alternate with third luminance pixel capturing elements; and a processing mechanism for transforming the first and second color components into full resolution color components and for transforming the luminance components into a 2X full resolution luminance component.
An object of the invention is provide a scanner which uses color interleaving with three one-dimensional sensors, where one sensor captures only the luminance, Y, and the
Carter Tia
Coles Edward
Robert D. Varitz, P.C
Sharp Laboratories of America Inc.
LandOfFree
Super resolution scanning using color multiplexing of 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 Super resolution scanning using color multiplexing of image..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Super resolution scanning using color multiplexing of image... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2970418