Document registration and skew detection system
View Patent ↗Disclosed methods and systems perform electronic registration of digitally captured images in real-time and performs accurate and robust digital image processing by analyzing the entire digitally captured image.
1. A digital imaging system, comprising:
a digital image capture device configured to generate a scan image that digitally represents an input document positioned in a scan area;
an image processor configured to receive and process said scan image in real-time, wherein said image processor includes an input document edge detection system, an input document corner detection system, an input document skew calculation system and an input document shear calculation system; and
an output processor configured to provide registration information for said input document,
said input document edge detection system further comprising:
a slow scan processor electronically linked to a lead edge detector and a trail edge detector, said lead edge detector being configured to detect a scan image scanline corresponding to an input document lead edge and said trail edge detector being configured to detect a scan image scanline corresponding to an input document trail edge; and
a fast scan processor electronically linked to a side edge detector and configured to detect a scan image scanline corresponding to an input document left edge and to detect a scan image scanline corresponding to an input document right edge.
2. A digital imaging system as claimed in claim 1 , wherein said registration information relates to an input document skew amount.
3. A digital imaging system as claimed in claim 1 , wherein said registration information relates to an input document shear amount.
4. A digital imaging system as claimed in claim 1 , wherein said registration information relates to an input document paper size.
5. A digital imaging system, comprising:
a digital image capture device configured to generate a scan image that digitally represents an input document positioned in a scan area;
an image processor configured to receive and process said scan image in real-time, wherein said image processor includes an input document edge detection system, an input document corner detection system, an input document skew calculation system and an input document shear calculation system; and
an output processor configured to provide registration information for said input document,
said input document corner detection system further comprising a corner pixel detector electronically linked to said fast scan processor and configured to identify pixels of said scan image that correspond to an input document corner, said input document corner pixel detection being independent of an output of said side edge detector.
6. A digital imaging system, comprising:
a digital image capture device configured to generate a scan image that digitally represents an input document positioned in a scan area;
an image processor configured to receive and process said scan image in real-time, wherein said image processor includes an input document edge detection system, an input document corner detection system, an input document skew calculation system and an input document shear calculation system; and
an output processor configured to provide registration information for said input document;
said input document skew calculation system further comprising:
a line calculator configured to calculate a skew angle and shear amount for pixels of said scan image that correspond to said input document based upon an output of said edge detection system; and
a corner processor configured to locate at least one pixel of said scan image that corresponds to a corner of said input document based upon said edge detection system output.
7. A digital imaging system, comprising:
a digital image capture device configured to generate a scan image that digitally represents an input document positioned in a scan area;
an image processor configured to receive and process said scan image in real-time, wherein said image processor includes an input document edge detection system, an input document corner detection system, an input document skew calculation system and an input document shear calculation system; and
an output processor configured to provide registration information for said input document;
a paper start detector electronically linked to said input document edge detection system and configured to identify pixels of said scan image that correspond to a first captured portion of said input document; and
a detection range controller electronically linked to said paper start detector and configured to limit an image processing operation to a predefined portion of said scan image.
8. A method, comprising:
Receiving a scan image that digitally represents an input document positioned in a scan area; and
operating an image processor to detect a fast-scan direction aligned input document edge by (i) obtaining a first grayscale average for a scan image target pixel and a pixel aligned with said target pixel in a slow-scan direction in a first previously processed scanline, (ii) obtaining a second grayscale average for pixels that are aligned with said target pixel in a slow-scan direction in at least a second and third previously processed scanline, (iii) obtaining a difference between said first grayscale average and said second grayscale average; (iv) obtaining a grayscale difference threshold that distinguishes background pixels and input document pixels and (v) designating said target pixel as a slow-scan transition pixel if said first and second grayscale average difference exceeds said grayscale difference threshold.
9. A method as claimed in claim 8 further comprising:
detecting a slow-scan direction aligned input document edge by (i) identifying a pixel segment of a scan image scanline; (ii) obtaining an average grayscale value for said scanline pixel segment; (iii) obtaining a grayscale value difference for said scanline pixel segment; (iv) obtaining a grayscale difference threshold that identifies a background-to-medium pixel; (v) obtaining a grayscale difference threshold that identifies a medium-to-background pixel; (vi) comparing said pixel segment average grayscale value to an input document threshold; (vii) comparing said scanline pixel segment average to a background image threshold; and (viii) identifying a first pixel segment in said scanline with a grayscale value average between said input document threshold and said background image threshold as a background-to-medium transition pixel; and (ix) identifying a last pixel segment in said scanline with a grayscale value average between said input document threshold and said background image threshold as a medium-to-background transition pixel.
10. A method as claimed in claim 9 further comprising:
comparing said scanline pixel segment difference to a background-to-medium threshold; comparing said scanline pixel segment difference to a medium-to-background threshold; and forwarding an output from said background-to-medium threshold comparison output and said medium-to-background threshold comparison output to a fast scan direction post processor for final determination of said left and right edge locations background-to-medium threshold.
11. A method as claimed in claim 10 further comprising entering grayscale difference values for adjacent pixels of a scanline into an array; and sorting said array to identify a scanline maximum grayscale value and a scanline minimum grayscale value.
12. A method as claimed in claim 8 further comprising detecting scan image scanlines corresponding to at least a lead edge and a trail edge of said input document.