Radiant energy – Invisible radiant energy responsive electric signalling – Infrared responsive
Reexamination Certificate
1998-01-23
2002-04-30
Sugarman, Scott J. (Department: 2873)
Radiant energy
Invisible radiant energy responsive electric signalling
Infrared responsive
C250S341100
Reexamination Certificate
active
06380539
ABSTRACT:
FIELD OF THE INVENTION
This invention relates to image enhancement and recovery, and more particularly to a method and apparatus for scanning color image data.
BACKGROUND OF THE INVENTION
Ever since the first image of an object was captured on film, a serious problem was apparent which has continued to plague the field of image capture and reproduction to the present day, namely imperfections in the recording medium itself which distort and obscure the original image sought to be captured. These imperfections occur in innumerable forms including dust, scratches, fingerprints, smudges and the like. Archival polypropylene sleeves employed to protect negatives even contribute to the problem by leaving hairline surface scratches as the negatives are pulled out of and replaced into the sleeves.
One method of approaching the problem of defective images is described by the present inventor in U.S. Pat. No. 5,266,805 issued to Edgar. In that system, image data is stored on a recording medium or substrate containing non-image imperfections such as film having surface scratches, wave smudges, bubbles, or the like, which give rise to undesirable artifacts in images subsequently retrieved from the medium or substrate. Means are provided for deriving from the medium separate images in the red, green, blue and infrared portions of the electromagnetic spectrum corresponding to the image stored. The infrared image is used as an indicator or map of the spatial position of the non-image imperfections on and in the medium so the defects can be reduced or eliminated. In this way, the desired underlying image is recovered.
The fundamentals of scanner technology for scanning and storing images in digital form are well developed and available in commercial products. Scanners receive an optical image, and divide it into many points, called pixels. The scanner measures light at each of these points to give each pixel a numerical value. Software associated with the scanner can then manipulate and store the pixel data. A color scanner measures several numerical values for each point, one for each primary color of red, green and blue. For example, a scanner may measure a pink pixel to have 80% red, 40% green, and 40% blue. All numbers representing one of the primary colors are grouped by the software into a channel, or single color image. Some scanners measure only one primary color at a time, then back up, change either the light source or a color filter to a second color, and measure only that second color on a second pass of the image. These are called multipass scanners. An example of a multipass scanner is the RFS 3570 made by Eastman Kodak Company.
Other scanners make a single pass, and collect all color information in that one pass. One type of single pass color scanner uses a linear array, or line of sensors. This array is moved, or scanned, perpendicularly across the image to scan all points on the image line by line. During this scan a light source is rapidly switched between the various colors to be sensed. One complete cycle of the colored light sources occurs for each line of pixels in the image. Thus a single point on the image is seen by a single sensor in the array first by one color, then by another color, and typically also by a third or fourth color, before that sensor moves on to another line of pixels on the image. Software is used to store the color-specific numerical information for each pixel into the appropriate channel so that at the end of the scan, multiple channels are available (one for each color). An example of a single pass scanner is the LS 1000 made by Nikon Corporation.
Another type of single pass scanner illuminates the image being scanned with light containing all visible colors, but places tiny color filters over the sensor elements so at any point in time there are sensors receiving each different color. One such method positions three linear arrays side by side to form a “trilinear array.” One of the three arrays is placed under a red filter, one under a green filter, and the third under a blue filter. Typically these colored filters pass infrared light in addition to the color they are intended to pass, and for this reason the infrared light must be removed from the optical path by a separate filter elsewhere in the scanner, or the scanner must use a light source containing no infrared light. An example of a scanner employing a trilinear array is the SprintScan 35 made by Polaroid Corporation.
A problem arises when applying media surface defect correction to single pass scanners using trilinear arrays. The filters on standard trilinear arrays distinguish red, green, and blue light, but not the fourth “color” of infrared light necessary to practice surface defect correction. Further, to be compatible with existing color dyes that pass infrared light, infrared light is often removed from the optical path prior to reaching the sensor, precluding addition of a fourth line of sensors sensitive to infrared light.
The three sensor lines of a trilinear array typically are spaced by an integer multiple of the separation distance between pixels in the image. If the spacing integer is eight, then a specific point on the image may be sensed in red, and exactly eight steps later it may be sensed in green, and eight steps after that it may be sensed in blue. Thus, the same pixel is scanned at three different times. The software then realigns the color channels by moving each a multiple of eight steps so as to group together all of the pixels sensed in the same color. The spacing between sensor lines of the array is usually a power of two (such as 2, 4, 8, 16, 32, etc.) multiplied by the pixel spacing distance to allow the designer the option to choose submultiple resolutions. For example, if the spacing were {fraction (1/250)}th of an inch, the scanner could operate at 2,000; 1,000; 500; or 250 dots per inch, while retaining alignment between the colors with the appropriate offset.
Trilinear array scanners operate at very high speed because three lines of an image are scanned simultaneously. In addition, they provide very good color registration because of the single pass and in spite of generally low cost transport mechanics. Thus, there is no need to halt movement of the image at each scan line, and this further increases the speed of the scan. However, the market needs an image scanner with the speed and cost advantages of a single pass trilinear array which also includes surface defect correction capabilities.
Yet another type of single pass scanner uses an area array to cover a two dimensional region of the image at once rather than mechanically scanning with a linear array to cover the region. One such scanner called a color filter matrix further incorporates tiny color filters over each element in the area array. In a specific implementation used in a Kodak digital camera, half the sensor elements lie under tiny green filters incorporated on the sensor array in a checkerboard pattern. The other half of the sensor elements in the checkerboard are behind alternating red and blue filters. Thus, a quarter of the sensors respond to red light, half respond to green light and a quarter respond to blue light. In another implementation in a Polaroid digital camera, an entire column of sensors is behind tiny red filters, the adjacent column of sensors is behind tiny green filters, and the next column of sensors is behind tiny blue filters. This pattern repeats in additional columns. In yet another implementation common in video cameras, the colored filters on even rows are green and magenta, and the filters on odd rows are cyan and yellow. Because the array is not mechanically scanned, each pixel in the image is measured with only one of the three colors, as opposed to the other scanners discussed so far that measure each pixel of the image with three colors. As with the trilinear array discussed above, all the colored filters pass infrared light, and therefore infrared light must be removed by a separate filter elsewhere in the optical path.
Another important consideration for
Applied Science Fiction Inc.
Dinsmore & Shohl LLP
Hanig Richard
Sugarman Scott J.
LandOfFree
Four color trilinear CCD scanning does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Four color trilinear CCD scanning, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Four color trilinear CCD scanning will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2836618