Back to results
US 6,785,416 B1Granted
System and method for the processing of scanned image data using a pixel window
AssigneeOak Technology, Inc.
InventorsShenbo Yu, Fred Andree, Steve Pratt, Wei Zhang
Life of this patent
Estimating the term…
Abstract
The illustrated embodiment processes scanned image data with a resolution of 600 dpi or higher. Image detection algorithms are used to detect the probable presence of halftone image data, large text image data, small text image data and photographic image data. The image detection algorithms are designed specifically for the amount of data contained in an image with a resolution of 600 dpi or greater. Once an image has been analyzed, the illustrated embodiment provides the correspondingly appropriate image process to the scanned image.
Claims36
A method of processing image data containing pixel values for pixels wherein said image data has a resolution of at least 600 dpi, said method comprising the steps of:
- positioning a pixel window over a subset of pixels in the image data, said pixel window surrounding a focus pixel;
- and performing calculations on pixel values of pixels in the pixel window to detect image type, said calculations distinguishing between a region containing large text image data and a region containing halftone image data, and wherein the performing of calculations further comprises the steps of:
- summing the pixel values of a first, second, third and fourth subset of pixels in said pixel window to form a respective first, second, third and fourth cumulative pixel value;
- subtracting said third cumulative pixel value from the first cumulative pixel value and the fourth cumulative pixel value from the second cumulative pixel value to produce respective results, the larger of the respective results being designated as an overall result, and comparing said overall result to previously obtained results for a region known to contain halftone image data and a region known to contain large text image data, to predict a likelihood of the pixel being examined belonging to a halftone image region or large text image region.
1 more element
A method of processing image data containing pixel values for pixels wherein said image data has a resolution of at least 600 dpi, said method comprising the steps of:
- positioning a pixel window over a subset of pixels in the image data, said pixel window divided into four equal sized quadrants surrounding a focus pixel being examined in said pixel window, said quadrants being oriented so as to have two upper quadrants and two lower quadrants and each upper quadrant having a corresponding lower quadrant that is diagonally positioned relative to the upper quadrant;
- performing calculations on pixel values of pixels in the pixel window to detect whether the focus pixel in the pixel window is in a region containing large text image data or a region containing halftone image data, and wherein the performing of calculations further comprises the steps of:
- summing the pixel values for the pixels in each said quadrant to form a cumulative quadrant pixel value;
- and
- subtracting said cumulative quadrant pixel value for each lower quadrant from the cumulative quadrant pixel values of the diagonally positioned upper quadrant to produce respective results, and designating the larger of the results as the result of said image detection algorithm, and
- comparing said result to previously obtained results for a region known to contain halftone image data and a region known to contain large text image data, to predict a likelihood of the pixel being examined belonging to a halftone image region or large text image region.
3 more elements
A method of processing image data containing pixel values for pixels wherein said image data has a resolution of at least 600 dpi, said method comprising the steps of:
- positioning a pixel window over a subset of pixels in the image data, said pixel window surrounding a focus pixel;
- and performing calculations on pixel values of pixels in the pixel window to detect image type, said calculations distinguishing between a region containing small text image data and a halftone image region, comparing the pixel values of the corner pixels of said pixel window to the value of said focus pixel so as to compare differences of pixel values between the focus pixel of the pixel window and the corner pixels at the corners of said pixel window to produce a first result;
- and comparing corner pixel values for pixels positioned in corners of a smaller pixel window, said smaller pixel window forming a subset of the first pixel window, to the value of said focus pixel to compare differences in pixel values between the focus pixel and the corner pixels at the corners of said slightly smaller pixel window to produce a second result;
- and comparing said first result and said second result to previously obtained results for a region known to contain small text image data, to predict a likelihood of the pixel being examined belonging to a small text image region.
1 more element
A method of processing image data containing pixel values for pixels wherein said image data has a resolution of at least 600 dpi, said method comprising the steps of:
- positioning a pixel window which surrounds a focus pixel over a subset of pixels in the image data, said pixel window being having four equal length sides joined together, a top side, a bottom side, a right side and a left side, corner pixels being located at the intersection of said top side and said right side, the intersection of said top side and said left side, the intersection of said bottom side and said left side, and the intersection of said bottom side and said right side;
- performing calculations on pixel values of pixels in the pixel window to detect whether the focus pixel in the pixel window is in a region containing small text image data, and wherein the performing of calculations further comprises the steps of:
- comparing the pixel values of the corner pixels of said pixel window to the value of said focus pixel so as to compare differences of pixel values between the focus pixel of the pixel window and the corner pixels at the corners of said pixel window to produce a first result;
- and
- comparing corner pixel values for pixels positioned in corners of a slightly smaller pixel window, forming a subset of the first pixel window to the value of said focus pixel to compare differences in pixel values between the focus pixel and the corner pixels at the corners of said slightly smaller pixel window to produce a second result;
- and
- comparing said first result and said second result to previously obtained results for a region known to contain small text image data, to predict a likelihood of the pixel being examined belonging to a small text image region.
4 more elements
A method of processing image data containing pixel values for pixels wherein said image data has a resolution of at least 600 dpi, said method comprising the steps of:
- positioning a pixel window over a subset of pixels in the image data, said pixel window surrounding a focus pixel;
- and
- performing calculations on pixel values of pixels in the pixel window to detect whether the focus pixel in the pixel window is in a region containing the text character ‘×’ or a region containing the text character ‘=’.
A method of processing image data containing pixel values for pixels wherein said image data has a resolution of at least 600 dpi, said method comprising the steps of:
- positioning a pixel window over a subset of pixels in the image data, said pixel window surrounding a focus pixel;
- and
- performing calculations on pixel values of pixels in the pixel window to detect whether the focus pixel in the pixel window is in a region containing the text character ‘≡’.
A method of processing image data containing pixel values for pixels wherein said image data has a resolution of at least 600 dpi, said method comprising the steps of:
- positioning a pixel window surrounding a focus pixel over a subset of pixels in the image, said pixel window being composed of horizontally oriented rows and vertically oriented columns of pixels, said pixel window containing pixels horizontally oriented relative to the focus pixel, said focus pixel also containing pixels vertically oriented to the focus pixel, said pixel window further containing pixels diagonally oriented to the focus pixel, said pixels forming a first diagonal relative to the focus pixel and a second diagonal relative to the focus pixel;
- performing calculations on pixel values of pixels in the pixel window to detect whether the focus pixel in the pixel window is in a region containing photographic image data, and wherein the step of performing calculations further comprises the steps of:
- comparing pixel values of a subset of the pixels horizontally oriented relative to said focus pixel to the pixel value of said focus pixel so as to compare differences of pixel values between the focus pixel of the pixel window and the subset of pixels horizontally oriented to form a first result;
- and
- comparing pixel values of a subset of the pixels vertically oriented relative to said focus pixel to the pixel value of said focus pixel so as to compare differences of pixel values between the focus pixel of the pixel window and the subset of pixels vertically oriented to form a second result;
- and
- comparing pixel values of a first subset of the pixels diagonally oriented relative to said focus pixel to the pixel value of said focus pixel so as to compare differences of pixel values between the focus pixel of the pixel window and the first subset of pixels diagonally oriented to form a third result;
- and
- comparing pixel values of a second subset of the pixels diagonally oriented relative to said focus pixel to the pixel value of said focus pixel so as to compare differences of pixel values between the focus pixel of the pixel window and the second subset of pixels diagonally oriented pixels to form a fourth result;
- and
- comparing said first, second, third and fourth results to previously obtained results for a region known to contain photographic image data, to predict a likelihood of the focus pixel belonging to a region containing photographic image data.
8 more elements
A method of processing image data containing pixel values for pixels wherein said image data has a resolution of 600 dpi, said method comprising the steps of:
- positioning a pixel window surrounding a pixel over a subset of pixels in the image data, said pixel window surrounding a focus pixel;
- and
- performing calculations on pixel values of pixels in the pixel window to detect whether the focus pixel in the pixel window is in a region containing large text image data, or a region containing halftone image data, or in a region containing small text image data, or in a region containing photographic image data, wherein the performing of calculations generates results which are then compared to previously obtained results for a known region of halftone image data, compared to results for a known region of large text image image data, compared to results for a known region of small text image data, and compared to results for a known area of photographic image, to predict a likelihood of the pixel being examined belonging to a halftone image region or large text image region, a small text image region or a photographic image region.
A medium holding computer-executable instructions for performing a method of processing image data containing pixel values for pixels wherein said image data has a resolution of 600 dpi, said method comprising the steps of:
- positioning a pixel window over a subset of pixels in the image data, said pixel window surrounding a focus pixel;
- and
- performing calculations on pixel values of pixels in the pixel window to detect whether the focus pixel in the pixel window is in a region containing large text image data, or in a region containing halftone image data, or in a region containing small text image data, or in a region containing photographic image data,
- and wherein the performing said calculations generates results which are then compared to previously obtained results for a known region of halftone image data, a known region of large text image data, a known region of small text image data, and a known region of photographic image data, to predict a likelihood of the focus pixel belonging to a halftone image region, or to a large text image region, or to a small text image region, or to a photographic image region.
1 more element
In an electronic device, a medium holding executable steps for a method of processing image data containing pixel values for pixels wherein said image data has a resolution of at least 600 dpi, said method comprising the steps of:
- positioning a pixel window over a subset of pixels in the image data, said pixel window divided into four equal sized quadrants surrounding a focus pixel being examined in said pixel window, said quadrants being oriented so as to have two upper quadrants and two lower quadrants and each upper quadrant having a corresponding lower quadrant that is diagonally positioned relative to the upper quadrant;
- performing calculations on pixel values of pixels in the pixel window to detect whether the focus pixel in the pixel window is in a region containing large text image data or a region containing halftone image data, and wherein the performing of calculations further comprises the steps of:
- summing the pixel values for the pixels in each said quadrant to form a cumulative quadrant pixel value;
- and
- subtracting said cumulative quadrant pixel value for each lower quadrant from the cumulative quadrant pixel values of the diagonally positioned upper quadrant to produce respective results, and designating the larger of the results as the result of said image detection algorithm, and
- comparing said result to previously obtained results for a region known to contain halftone image data and a region known to contain large text image data, to predict a likelihood of the pixel being examined belonging to a halftone image region or large text image region.
3 more elements
In an electronic device, a medium holding executable steps for a method of processing image data containing pixel values for pixels wherein said image data has a resolution of at least 600 dpi, said method comprising the steps of:
- positioning a pixel window which surrounds a focus pixel over a subset of pixels in the image data, said pixel window being having four equal length sides joined together, a top side, a bottom side, a right side and a left side, corner pixels being located at the intersection of said top side and said right side, the intersection of said top side and said left side, the intersection of said bottom side and said left side, and the intersection of said bottom side and said right side;
- performing calculations on pixel values of pixels in the pixel window to detect whether the focus pixel in the pixel window is in a region containing small text image data, and wherein the performing of calculations further comprises the steps of:
- comparing the pixel values of the corner pixels of said pixel window to the value of said focus pixel so as to compare differences of pixel values between the focus pixel of the pixel window and the corner pixels at the corners of said pixel window to produce a first result;
- and
- comparing corner pixel values for pixels positioned in corners of a slightly smaller pixel window, forming a subset of the first pixel window to the value of said focus pixel to compare differences in pixel values between the focus pixel and the corner pixels at the corners of said slightly smaller pixel window to produce a second result;
- and
- comparing said first result and said second result to previously obtained results for a region known to contain small text image data, to predict a likelihood of the pixel being examined belonging to a small text image region.
4 more elements
In an electronic device, a medium holding executable steps for a method of processing image data containing pixel values for pixels wherein said image data has a resolution of at least 600 dpi, said method comprising the steps of:
- positioning a pixel window surrounding a focus pixel over a subset of pixels in the image, said pixel window being composed of horizontally oriented rows and vertically oriented columns of pixels, said pixel window containing pixels horizontally oriented relative to the focus pixel, said focus pixel also containing pixels vertically oriented to the focus pixel, said pixel window further containing pixels diagonally oriented to the focus pixel, said pixels forming a first diagonal relative to the focus pixel and a second diagonal relative to the focus pixel;
- performing calculations on pixel values of pixels in the pixel window to detect whether the focus pixel in the pixel window is in a region containing photographic image data, and wherein the step of performing calculations further comprises the steps of:
- comparing pixel values of a subset of the pixels horizontally oriented relative to said focus pixel to the pixel value of said focus pixel so as to compare differences of pixel values between the focus pixel of the pixel window and the subset of pixels horizontally oriented to form a first result;
- and
- comparing pixel values of a subset of the pixels vertically oriented relative to said focus pixel to the pixel value of said focus pixel so as to compare differences of pixel values between the focus pixel of the pixel window and the subset of pixels vertically oriented to form a second result;
- and
- comparing pixel values of a first subset of the pixels diagonally oriented relative to said focus pixel to the pixel value of said focus pixel so as to compare differences of pixel values between the focus pixel of the pixel window and the first subset of pixels diagonally oriented to form a third result;
- and
- comparing pixel values of a second subset of the pixels diagonally oriented relative to said focus pixel to the pixel value of said focus pixel so as to compare differences of pixel values between the focus pixel of the pixel window and the second subset of pixels diagonally oriented pixels to form a fourth result;
- and
- comparing said first, second, third and fourth results to previously obtained results for a region known to contain photographic image data, to predict a likelihood of the focus pixel belonging to a region containing photographic image data.
8 more elements
In an electronic device, a medium holding executable steps for a method of processing image data containing pixel values for pixels wherein said image data has a resolution of at least 600 dpi, said method comprising the steps of:
- positioning a pixel window over a subset of pixels in the image data, said pixel window surrounding a focus pixel;
- and
- performing calculations on pixel values of pixels in the pixel window to detect whether the focus pixel in the pixel window is in a region containing the text character ‘×’ or a region containing the text character ‘=’.
In an electronic device, a medium holding executable steps for a method of processing image data containing pixel values for pixels wherein said image data has a resolution of at least 600 dpi, said method comprising the steps of:
- positioning a pixel window over a subset of pixels in the image data, said pixel window surrounding a focus pixel;
- and
- performing calculations on pixel values of pixels in the pixel window to detect whether the focus pixel in the pixel window is in a region containing the text character ‘≡’.
Dependencies are inferred from the claim text.