Image enhancement system

Claim 2

A digital image processing apparatus comprising in combination; A. means for receiving a scaler image file of scaled pixel values; B. means for segmentation of the scaled pixel values into graphics blocks and non-graphics blocks, the segmentation means including: 1. means for determining at least a first edge threshold from the scaled pixel values wherein the edge threshold is determined according to the following formula: TV=a×((LNG-DNG)/b+DNG); where: a and b are numerical coefficients; TV=edge threshold; LNG=lightest non-white scalar pixel value; and DNG=darkest non-black scalar pixel value; and 2. means for applying at least one edge detection filter to each scalar pixel value and tagging as an edge the scalar pixel values satisfying the conditions of the edge detection filter; C. means for developing a binary image file derived from the scalar image file, the developing means including: 1. means for applying a text threshold to the non-graphics blocks in the scalar image file to convert each scaled pixel non-graphic value into a first binary value representing text or alternatively a second binary value representing background for the text; and 2. means for applying a graphics threshold to graphics blocks in the scalar image file to convert each scaled pixel graphic value into a first binary value representing one color or alternatively a second binary value representing another color; and D. means for compressing the binary image file to yield a compressed image file for subsequent decompression and digital processing of the compressed image file.

Claim 3

The digital image processing apparatus of claim 1 wherein the segmentation means also includes block division and tagging means having: (i) means for locating blocks of cells within a predetermined proximity to one another whose contents have been tagged as text; (ii) means for analyzing cells in the block and re-tagging them as text or graphics; and (iii) means for comparing any remaining cells, not previously allocated to blocks, to adjacent cells to re-tag the remaining cells as text or graphics.

Claim 4

The digital image processing apparatus of claim 2 also including a compressor whereby the binary image file is compressed to yield a compressed image file for subsequent decompression and digital processing of the compressed image file and wherein the compressor implements the CCITT Group 3 or 4 standard.

Claim 5

The digital image processing apparatus of claim 3 also including means for deskewing the image.

Claim 6

The digital image processing apparatus of claim 4 also including: (i) means for filtration of the text blocks to clarify the textual components represented by the text blocks; and (ii) means for filtration of the graphics blocks to enhance the graphics components represented by the graphics blocks.

Claim 7

The digital image processing apparatus of claim 1 also including means for deskewing the image.

Claim 8

The digital image processing apparatus of claim 6 wherein the deskewing means includes: A. means for tracing the image file to locate a rectangle boundary for the image file; B. means for computing the widest points of the rectangle boundary; C. means for determining a center for centering the four corners of the image file; D. means for locating the longest straight line on at least one side of the rectangle boundary; E. means for computation of the skew angle for at least the one side from optimal horizontal or vertical alignment about the center; and F. means for applying a two-dimensional rotation of the rectangle boundary about the center through the skew angle.

Claim 9

A digital image processing apparatus comprising in combination: A. means for receiving a scaler image file of scaled pixel value; B. means for segmentation of the scaled pixel values into graphics blocks and non-graphics blocks, the segmentation means including: 1. means for determining at least a first edge threshold from the scaled pixel values wherein the edge threshold is determined according to the following formula: TV=a×((LNG-DNG)/b+DNG); where: a and b are numerical coefficients; TV=edge threshold; LNG=lightest non-white scalar pixel value; and DNG=darkest non-black scalar pixel value; and 2. means for applying at least one edge detection filter to each scalar pixel value and tagging as an edge the scalar pixel values satisfying the conditions of the edge detection filter; C. means for developing a binary image file derived from the scalar image file, the developing means including: 1. means for applying a text threshold to the non-graphics blocks in the scalar image file to convert each scaled pixel non-graphic value into a first binary value representing text or alternatively a second binary value representing background for the text; the applying means includes a vertical edge detection filter and a horizontal edge detection filter with the vertical edge detection filter emphasized in comparison to the horizontal edge detection filter so that the applying means is more likely to locate a vertical edge than a horizontal edge; and 2. means for applying a graphics threshold to graphics blocks in the scalar image file to convert each scaled pixel graphic value into a first binary value representing one color or alternatively a second binary value representing another color; and D. means for compressing the binary image file to yield a compressed image file for subsequent decompression and digital processing of the compressed image file.

Claim 10

A digital image processing apparatus comprising in combination: A. means for receiving a scaler image file of scaled pixel value; B. means for segmentation of the scaled pixel values into graphics blocks and non-graphics blocks, the segmentation means including: 1. means for determining at least a first edge threshold from the scaled pixel values wherein the edge threshold is determined according to the following formula: TV=a×((LNG-DNG)/b+DNG); where: a and b are numerical coefficients; TV=edge threshold; LNG=lightest non-white scalar pixel value; and DNG=darkest non-black scalar pixel value; and 2. means for applying at least one edge detection filter to each scalar pixel value and tagging as an edge the scalar pixel values satisfying the conditions of the edge detection filter; C. means for developing a binary image file derived from the scalar image file, the developing means including: 1. means for applying a text threshold to the non-graphics blocks in the scalar image file to convert each scaled pixel non-graphic value into a first binary value representing text or alternatively a second binary value representing background for the text; the applying means includes a vertical edge detection filter and a horizontal edge detection filter with the vertical edge detection filter emphasized in comparison to the horizontal edge detection filter so that the applying means is more likely to locate a vertical edge than a horizontal edge; and 2. means for applying a graphics threshold to graphics blocks in the scalar image file to convert each scaled pixel graphic value into a first binary value representing one color or alternatively a second binary value representing another color; and D. means for compressing the binary image file to yield a compressed image file for subsequent decompression and digital processing of the compressed image file.

Claim 11

The digital image processing apparatus of claim 8 also including: (i) means for filtration of the text blocks to clarify the textual components represented by the text blocks; and (ii) means for filtration of the graphics blocks to enhance the graphics components represented by the graphics blocks.

Claim 12

The digital image processing apparatus of claim 8 also including a compressor whereby the binary image file is compressed to yield a compressed image file for subsequent decompression and digital processing of the compressed image file and wherein the compressor implements the CCITT Group 3 or 4 standard.

Claim 13

The digital image processing apparatus of claim 8 also including means for deskewing the image.

Claim 14

The digital image processing apparatus of claim 11 wherein the deskewing means includes: A. means for tracing the image file to locate a rectangle boundary for the image file; B. means for computing the widest points of the rectangle boundary; C. means for determining a center for centering the four corners of the image file; D. means for locating the longest straight line on at least one side of the rectangle boundary; E. means for computation of the skew angle for at least the one side from optimal horizontal or vertical alignment about the center; and F. means for applying a two-dimensional rotation of the rectangle boundary about the center through the skew angle.

Claim 15

A digital image processing apparatus comprising in combination: A. means for receiving a scaler image file of scaled pixel values; B. means for segmentation of the scaled pixel values into graphics blocks and non-graphics blocks, the segmentation means including: 1. means for determining at least a first edge threshold from the scaled pixel values wherein the edge threshold is determined according to the following formula: TV=a×((LNG-DNG)/b+DNG); where: a and b are numerical coefficients; TV=edge threshold; LNG=lightest non-white scalar pixel value; and DNG=darkest non-black scalar pixel value; 2. means for applying at least one edge detection filter to each scalar pixel value and tagging as an edge the scalar pixel values satisfying the conditions of the edge detection filter; and 3. means for grouping adjacent scalar pixel values into a cell and re-tagging the scalar pixel values in the cell as: (a) text if the cell contains outer connected edge of adjacent scalar pixel values or an edge of scalar pixel values extending to the boundaries of the cell; (b) background if the cell contains a majority of non-edge scalar pixel values; and (c) graphics otherwise; C. means for developing a binary image file derived from the scalar image file, the developing means including: 1. means for applying a text threshold to the non-graphics blocks in the scalar image file to convert each scaled pixel non-graphic value into a first binary value representing text or alternatively a second binary value representing background for the text; and 2. means for applying a graphics threshold to graphics blocks in the scalar image file to convert each scaled pixel graphic value into a first binary value representing one color or alternatively a second binary value representing another color; and D. means for compressing the binary image file to yield a compressed image file for subsequent decompression and digital processing of the compressed image file.

Claim 16

A digital image processing apparatus comprising in combination: A. means for receiving a scaler image file of scaled pixel values; B. means for segmentation of the scaled pixel values into graphics blocks and non-graphics blocks, the segmentation means including: 1. means for determining at least a first edge threshold from the scaled pixel values wherein the edge threshold is determined according to the following formula: TV=a×((LNG-DNG)/b+DNG); where: a and b are numerical coefficients; TV=edge threshold; LNG=lightest non-white scalar pixel value; and DNG=darkest non-black scalar pixel value; 2. means for applying at least one edge detection filter to each scalar pixel value and tagging as an edge the scalar pixel values satisfying the conditions of the edge detection filter; and 3. means for grouping adjacent scalar pixel values into a cell and re-tagging the scalar pixel values in the cell as: (a) text if the cell contains outer connected edge of adjacent scalar pixel values or an edge of scalar pixel values extending to the boundaries of the cell; (b) background if the cell contains a majority of non-edge scalar pixel values; and (c) graphics otherwise; C. means for developing a binary image file derived from the scalar image file, the developing means including: 1. means for applying a text threshold to the non-graphics blocks in the scalar image file to convert each scaled pixel non-graphic value into a first binary value representing text or alternatively a second binary value representing background for the text; and 2. means for applying a graphics threshold to graphics blocks in the scalar image file to convert each scaled pixel graphic value into a first binary value representing one color or alternatively a second binary value representing another color; and D. means for compressing the binary image file to yield a compressed image file for subsequent decompression and digital processing of the compressed image file.

Claim 17

The digital image processing apparatus of claim 13 also including: (i) means for filtration of the text blocks to clarify the textual components represented by the text blocks; and (ii) means for filtration of the graphics blocks to enhance the graphics components represented by the graphics blocks.

Claim 18

The digital image processing apparatus of claim 13 also including a compressor whereby the binary image file is compressed to yield a compressed image file for subsequent decompression and digital processing of the compressed image file and wherein the compressor implements the CCITT Group 3 or 4 standard.

Claim 19

The digital image processing apparatus of claim 13 also including means for deskewing the image.

Claim 20

The digital image processing apparatus of claim 16 wherein the deskewing means includes: A. means for tracing the image file to locate a rectangle boundary for the image file; B. means for computing the widest points of the rectangle boundary; C. means for determining a center for centering the four corners of the image file; D. means for locating the longest straight line on at least one side of the rectangle boundary; E. means for computation of the skew angle for at least the one side from optimal horizontal or vertical alignment about the center; and F. means for applying a two-dimensional rotation of the rectangle boundary about the center through the skew angle.

Claim 21

A digital image processing apparatus comprising in combination: A. means for receiving a scaler image file of scaled pixel value; B. means for segmentation of the scaled pixel values into graphics blocks and non-graphics blocks, the segmentation means including: 1. means for determining at least a first edge threshold from the scaled pixel values wherein the edge threshold is determined according to the following formula: TV=a×((LNG-DNG)/b+DNG); where: a and b are numerical coefficients; TV=edge threshold; LNG=lightest non-white scalar pixel value; and DNG=darkest non-black scalar pixel value; and 2. means for applying at least one edge detection filter to each scalar pixel value and tagging as an edge the scalar pixel values satisfying the conditions of the edge detection filter; C. means for developing a binary image file derived from the scalar image file, the developing means including: 1. means for applying a text threshold to the non-graphics blocks in the scalar image file to convert each scaled pixel non-graphic value into a first binary value representing text or alternatively a second binary value representing background for the text; the applying means includes a vertical edge detection filter and a horizontal edge detection filter with the vertical edge detection filter emphasized in comparison to the horizontal edge detection filter so that the applying means is more likely to locate a vertical edge than a horizontal edge; and 2. means for applying a graphics threshold to graphics blocks in the scalar image file to convert each scaled pixel graphic value into a first binary value representing one color or alternatively a second binary value representing another color; and D. means for compressing the binary image file to yield a compressed image file for subsequent decompression and digital processing of the compressed image file.

Claim 22

A digital image processing apparatus comprising in combination: A. means for receiving a scaler image file of scaled pixel value; B. means for segmentation of the scaled pixel values into graphics blocks and non-graphics blocks, the segmentation means including: 1. means for determining at least a first edge threshold from the scaled pixel values wherein the edge threshold is determined according to the following formula: TV=a×((LNG-DNG)/b+DNG); where: a and b are numerical coefficients; TV=edge threshold; LNG=lightest non-white scalar pixel value; and DNG=darkest non-black scalar pixel value; and 2. means for applying at least one edge detection filter to each scalar pixel value and tagging as an edge the scalar pixel values satisfying the conditions of the edge detection filter; C. means for developing a binary image file derived from the scalar image file, the developing means including: 1. means for applying a text threshold to the non-graphics blocks in the scalar image file to convert each scaled pixel non-graphic value into a first binary value representing text or alternatively a second binary value representing background for the text; the applying means includes a vertical edge detection filter and a horizontal edge detection filter with the vertical edge detection filter emphasized in comparison to the horizontal edge detection filter so that the applying means is more likely to locate a vertical edge than a horizontal edge; and 2. means for applying a graphics threshold to graphics blocks in the scalar image file to convert each scaled pixel graphic value into a first binary value representing one color or alternatively a second binary value representing another color; and D. means for compressing the binary image file to yield a compressed image file for subsequent decompression and digital processing of the compressed image file.

Claim 23

The digital image processing apparatus of claim 18 also including means for deskewing the image.

Claim 24

The digital image processing apparatus of claim 19 wherein the deskewing means includes: A. means for tracing the image file to locate a rectangle boundary for the image file; B. means for computing the widest points of the rectangle boundary; C. means for determining a center for centering the four corners of the image file; D. means for locating the longest straight line on at least one side of the rectangle boundary; E. means for computation of the skew angle for at least the one side from optimal horizontal or vertical alignment about the center; and F. means for applying a two-dimensional rotation of the rectangle boundary about the center through the skew angle.

Claim 25

A digital image processing apparatus comprising in combination: A. means for receiving a scaler image file of scaled pixel value; B. means for segmentation of the scaled pixel values into graphics blocks and non-graphics blocks, the segmentation means including: 1. means for determining at least a first edge threshold from the scaled pixel values; 2. means for applying at least one edge detection filter to each scalar pixel value and tagging as an edge the scalar pixel values satisfying the conditions of the edge detection filter; and 3. block division and tagging means having: (i) means for locating blocks of cells within a predetermined proximity to one another whose contents have been tagged as text; (ii) means for analyzing cells in the block and re-tagging them as text or graphics wherein the analyzing means tags the cells in the block according to the following procedure, in which a-g are fractional coefficients: A. Define NUMCELLS=Number of cells in the block. B. If:

Claim 26

A digital image processing apparatus comprising in combination: A. means for receiving a scaler image file of scaled pixel value; B. means for segmentation of the scaled pixel values into graphics blocks and non-graphics blocks, the segmentation means including: 1. means for determining at least a first edge threshold from the scaled pixel values; 2. means for applying at least one edge detection filter to each scalar pixel value and tagging as an edge the scalar pixel values satisfying the conditions of the edge detection filter; and 3. block division and tagging means having: (i) means for locating blocks of cells within a predetermined proximity to one another whose contents have been tagged as text; (ii) means for analyzing cells in the block and re-tagging them as text or graphics wherein the analyzing means tags the cells in the block according to the following procedure, in which a-g are fractional coefficients: A. Define NUMCELLS=Number of cells in the block. B. If:

Claim 27

The number of graphics cells>=NUMCELLS a or: 2. both (i) the number of the text cells>=NUMCELLS * b, and (ii) the number of white cells=NUMCELLS * c; 3. Then tag all cells in the bock as containing graphics. C. If: 1. the number of text cells>=NUMCELLS * d; or 2. both (i) the number of text cells>=NUMCELLS * e, and (ii) the number of white cells>=NUMCELLS * f; or 3. the number of white cells>=NUMCELLS * g; 4. then tag the cells in the block as text. D. Otherwise: 1. subdivide the block into nine different regions. 2. Starting with the upper-left region, repeat the process of steps A through C above, treating the region as a block of cells in the context of those steps. 3. For any region that cannot be tagged according to the process of steps A through C above, treating the region as a block in the context of steps D(1) and D(2) above, and repeat this subdividing, testing, and tagging procedure until: a. all cells in the block are re-tagged; or b. the remaining sub-regions of cells that have not yet been re-tagged consist of ten or fewer cells. In this event, do not change the original tagging for these sub-regions; and (iii) means for analyzing cells in the block and re-tagging them as text or graphics; and (iv) means for comparing any remaining cells, not previously allocated to blocks, to adjacent cells to re-tag the remaining cells as text or graphics; C. means for developing a binary image file derived from the scalar image file, the developing means including:

Claim 28

The number of graphics cells>=NUMCELLS a or: 2. both (i) the number of the text cells>=NUMCELLS * b, and (ii) the number of white cells=NUMCELLS * c; 3. Then tag all cells in the bock as containing graphics. C. If: 1. the number of text cells>=NUMCELLS * d; or 2. both (i) the number of text cells>=NUMCELLS * e, and (ii) the number of white cells>=NUMCELLS * f; or 3. the number of white cells>=NUMCELLS * g; 4. then tag the cells in the block as text. D. Otherwise: 1. subdivide the block into nine different regions. 2. Starting with the upper-left region, repeat the process of steps A through C above, treating the region as a block of cells in the context of those steps. 3. For any region that cannot be tagged according to the process of steps A through C above, treating the region as a block in the context of steps D(1) and D(2) above, and repeat this subdividing, testing, and tagging procedure until: a. all cells in the block are re-tagged; or b. the remaining sub-regions of cells that have not yet been re-tagged consist of ten or fewer cells. In this event, do not change the original tagging for these sub-regions; and (iii) means for analyzing cells in the block and re-tagging them as text or graphics; and (iv) means for comparing any remaining cells, not previously allocated to blocks, to adjacent cells to re-tag the remaining cells as text or graphics; C. means for developing a binary image file derived from the scalar image file, the developing means including:

Claim 29

means for applying a text threshold to the non-graphics blocks in the scalar image file to convert each scaled pixel non-graphic value into a first binary value representing text or alternatively a second binary value representing background for the text: the applying means includes a vertical edge detection filter and a horizontal edge detection filter with the vertical edge detection filter emphasized in comparison to the horizontal edge detection filter so that the applying means is more likely to locate a vertical edge than a horizontal edge; and 2. means for applying a graphics threshold to graphics blocks in the scalar image file to convert each scaled pixel graphic value into a first binary value representing one color or alternatively a second binary value representing another color; and D. means for compressing the binary image file to yield a compressed image file for subsequent decompression and digital processing of the compressed image file.

Claim 30

means for applying a text threshold to the non-graphics blocks in the scalar image file to convert each scaled pixel non-graphic value into a first binary value representing text or alternatively a second binary value representing background for the text: the applying means includes a vertical edge detection filter and a horizontal edge detection filter with the vertical edge detection filter emphasized in comparison to the horizontal edge detection filter so that the applying means is more likely to locate a vertical edge than a horizontal edge; and 2. means for applying a graphics threshold to graphics blocks in the scalar image file to convert each scaled pixel graphic value into a first binary value representing one color or alternatively a second binary value representing another color; and D. means for compressing the binary image file to yield a compressed image file for subsequent decompression and digital processing of the compressed image file.

Claim 31

The digital image processing apparatus of claim 21 also including means for deskewing the image.

Claim 32

The digital image processing apparatus of claim 22 wherein the deskewing means includes: A. means for tracing the image file to locate a rectangle boundary for the image file; B. means for computing the widest points of the rectangle boundary; C. means for determining a center for centering the four corners of the image file; D. means for locating the longest straight line on at least one side of the rectangle boundary; E. means for computation of the skew angle for at least the one side from optimal horizontal or vertical alignment about the center; and F. means for applying a two-dimensional rotation of the rectangle boundary about the center through the skew angle.

Claim 33

A digital image processing apparatus comprising in combination: A. means for receiving a scaler image file of scaled pixel value; B. means for segmentation of the scaled pixel values into graphics blocks and non-graphics blocks, the segmentation means including: 1. means for determining at least a first edge threshold from the scaled pixel values; 2. means for applying at least one edge detection filter to each scalar pixel value and tagging as an edge the scalar pixel values satisfying the conditions of the edge detection filter; 3. means for grouping adjacent scalar pixel values into a cell and re-tagging the scalar pixel values in the cell as: (a) text if the cell contains outer connected edge of adjacent scalar pixel values or an edge of scalar pixel values extending to the boundaries of the cell; (b) background if the cell contains a majority of non-edge scalar pixel values; and (c) graphics otherwise; and 3. block division and tagging means having: (i) means for locating blocks of cells within a predetermined proximity to one another whose contents have been tagged as text; (ii) means for analyzing cells in the block and re-tagging them as text or graphics wherein the analyzing means tags the cells in the block according to the following procedure, in which a-g are fractional coefficients: A. Define NUMCELLS=Number of cells in the block. B. If:

Claim 34

A digital image processing apparatus comprising in combination: A. means for receiving a scaler image file of scaled pixel value; B. means for segmentation of the scaled pixel values into graphics blocks and non-graphics blocks, the segmentation means including: 1. means for determining at least a first edge threshold from the scaled pixel values; 2. means for applying at least one edge detection filter to each scalar pixel value and tagging as an edge the scalar pixel values satisfying the conditions of the edge detection filter; 3. means for grouping adjacent scalar pixel values into a cell and re-tagging the scalar pixel values in the cell as: (a) text if the cell contains outer connected edge of adjacent scalar pixel values or an edge of scalar pixel values extending to the boundaries of the cell; (b) background if the cell contains a majority of non-edge scalar pixel values; and (c) graphics otherwise; and 3. block division and tagging means having: (i) means for locating blocks of cells within a predetermined proximity to one another whose contents have been tagged as text; (ii) means for analyzing cells in the block and re-tagging them as text or graphics wherein the analyzing means tags the cells in the block according to the following procedure, in which a-g are fractional coefficients: A. Define NUMCELLS=Number of cells in the block. B. If:

Claim 35

The number of graphics cells>=NUMCELLS a or: both (i) the number of the text cells>=NUMCELLS * b, and (ii) the number of white cells=NUMCELLS * c; 3. Then tag all cells in the bock as containing graphics. C. If: 1. the number of text cells>=NUMCELLS * d; or 2. both (i) the number of text cells>=NUMCELLS * e, and (ii) the number of white cells>=NUMCELLS * f; or 3. the number of white cells>=NUMCELLS * g; 4. then tag the cells in the block as text. D. Otherwise: 1. subdivide the block into nine different regions. 2. Starting with the upper-left region, repeat the process of steps A through C above, treating the region as a block of cells in the context of those steps.

Claim 36

The number of graphics cells>=NUMCELLS a or: both (i) the number of the text cells>=NUMCELLS * b, and (ii) the number of white cells=NUMCELLS * c; 3. Then tag all cells in the bock as containing graphics. C. If: 1. the number of text cells>=NUMCELLS * d; or 2. both (i) the number of text cells>=NUMCELLS * e, and (ii) the number of white cells>=NUMCELLS * f; or 3. the number of white cells>=NUMCELLS * g; 4. then tag the cells in the block as text. D. Otherwise: 1. subdivide the block into nine different regions. 2. Starting with the upper-left region, repeat the process of steps A through C above, treating the region as a block of cells in the context of those steps.

Claim 37

For any region that cannot be tagged according to the process of steps A through C above, treating the region as a block in the context of steps D(1) and D(2) above, and repeat this subdividing, testing, and tagging procedure until: a. all cells in the block are re-tagged; or b. the remaining sub-regions of cells that have not yet been re-tagged consist of ten or fewer cells. In this event, do not change the original tagging for these sub-regions; and (iii) means for analyzing cells in the block and re-tagging them as text or graphics; and (iv) means for comparing any remaining cells, not previously allocated to blocks, to adjacent cells to re-tag the remaining cells as text or graphics; C. means for developing a binary image file derived from the scalar image file, the developing means including: 1. means for applying a text threshold to the non-graphics blocks in the scalar image file to convert each scaled pixel non-graphic value into a first binary value representing text or alternatively a second binary value representing background for the text; the applying means includes a vertical edge detection filter and a horizontal edge detection filter with the vertical edge detection filter emphasized in comparison to the horizontal edge detection filter so that the applying means is more likely to locate a vertical edge than a horizontal edge; and 2. means for applying a graphics threshold to graphics blocks in the scalar image file to convert each scaled pixel graphic value into a first binary value representing one color or alternatively a second binary value representing another color; and D. means for compressing the binary image file to yield a compressed image file for subsequent decompression and digital processing of the compressed image file.

Claim 38

For any region that cannot be tagged according to the process of steps A through C above, treating the region as a block in the context of steps D(1) and D(2) above, and repeat this subdividing, testing, and tagging procedure until: a. all cells in the block are re-tagged; or b. the remaining sub-regions of cells that have not yet been re-tagged consist of ten or fewer cells. In this event, do not change the original tagging for these sub-regions; and (iii) means for analyzing cells in the block and re-tagging them as text or graphics; and (iv) means for comparing any remaining cells, not previously allocated to blocks, to adjacent cells to re-tag the remaining cells as text or graphics; C. means for developing a binary image file derived from the scalar image file, the developing means including: 1. means for applying a text threshold to the non-graphics blocks in the scalar image file to convert each scaled pixel non-graphic value into a first binary value representing text or alternatively a second binary value representing background for the text; the applying means includes a vertical edge detection filter and a horizontal edge detection filter with the vertical edge detection filter emphasized in comparison to the horizontal edge detection filter so that the applying means is more likely to locate a vertical edge than a horizontal edge; and 2. means for applying a graphics threshold to graphics blocks in the scalar image file to convert each scaled pixel graphic value into a first binary value representing one color or alternatively a second binary value representing another color; and D. means for compressing the binary image file to yield a compressed image file for subsequent decompression and digital processing of the compressed image file.

Claim 39

A digital image processing apparatus comprising in combination: A. means for receiving a scaler image file of scaled pixel value; B. means for segmentation of the scaled pixel values into graphics blocks and non-graphics blocks; C. means for developing a binary image file derived from the scalar image file, the developing means including: 1. means for applying a text threshold to the non-graphics blocks in the scalar image file to convert each scaled pixel non-graphic value into a first binary value representing text or alternatively a second binary value representing background for the text; 2. means for applying a graphics threshold to graphics blocks in the scalar image file to convert each scaled pixel graphic value into a first binary value representing one color or alternatively a second binary value representing another color; and 3. means for subdividing the image into regions and for each such region calculating a text threshold or a graphics threshold from the scalar pixel values in the region; wherein the text and graphics thresholds are calculated according to the following formulas: TT=a * (MSV+((LNG-DNG)/b+DNG) GT=(LNG-DNG)* a+DNG+(d * ((USS-LSS)/b+USS))e/f where: a, b, c, d, e, and f are numerical coefficients; TT=text threshold; GT=graphics threshold; MSV=median scalar pixel value; LNG=lightest non-white scalar pixel value; and DNG=darkest non-black scalar pixel value; and D. means for compressing the binary image file to yield a compressed image file for subsequent decompression and digital processing of the compressed image file.

Claim 40

A digital image processing apparatus comprising in combination: A. means for receiving a scaler image file of scaled pixel value; B. means for segmentation of the scaled pixel values into graphics blocks and non-graphics blocks; C. means for developing a binary image file derived from the scalar image file, the developing means including: 1. means for applying a text threshold to the non-graphics blocks in the scalar image file to convert each scaled pixel non-graphic value into a first binary value representing text or alternatively a second binary value representing background for the text; 2. means for applying a graphics threshold to graphics blocks in the scalar image file to convert each scaled pixel graphic value into a first binary value representing one color or alternatively a second binary value representing another color; and 3. means for subdividing the image into regions and for each such region calculating a text threshold or a graphics threshold from the scalar pixel values in the region; wherein the text and graphics thresholds are calculated according to the following formulas: TT=a * (MSV+((LNG-DNG)/b+DNG) GT=(LNG-DNG)* a+DNG+(d * ((USS-LSS)/b+USS))e/f where: a, b, c, d, e, and f are numerical coefficients; TT=text threshold; GT=graphics threshold; MSV=median scalar pixel value; LNG=lightest non-white scalar pixel value; and DNG=darkest non-black scalar pixel value; and D. means for compressing the binary image file to yield a compressed image file for subsequent decompression and digital processing of the compressed image file.

Claim 41

A digital image processing apparatus comprising in combination: A. means for receiving a scaler image file of scaled pixel values; B. means for segmentation of the scaled pixel values into graphics blocks and non-graphics blocks; C. means for developing a binary image file derived from the scalar image file, the developing means including:

Claim 42

A digital image processing apparatus comprising in combination: A. means for receiving a scaler image file of scaled pixel values; B. means for segmentation of the scaled pixel values into graphics blocks and non-graphics blocks; C. means for developing a binary image file derived from the scalar image file, the developing means including:

Claim 43

means for applying a text threshold to the non-graphics blocks in the scalar image file to convert each scaled pixel non-graphic value into a first binary value representing text or alternatively a second binary value representing background for the text; and 2. means for applying a graphics threshold to graphics blocks in the scalar image file to convert each scaled pixel graphic value into a first binary value representing one color or alternatively a second binary value representing another color wherein the graphics threshold application means includes: (a) means for passing a dither matrix over the graphics blocks to re-assign a background pixel value to all scalar pixel values in the block less than a corresponding dither value in the dither matrix; (b) means for passing an error diffuse dither matrix over the graphics blocks to yield error diffuse pixel values; (c) comparing the graphics threshold value to the difference between the non-re-assigned scalar pixel values and the corresponding error diffuse pixel values; and (d) means for re-setting the scalar pixel value to a first binary number if the difference is greater than the graphics threshold and for otherwise re-setting the scalar pixel value to a second binary number; and D. means for compressing the binary image file to yield a compressed image file for subsequent decompression and digital processing of the compressed image file.

Claim 44

means for applying a text threshold to the non-graphics blocks in the scalar image file to convert each scaled pixel non-graphic value into a first binary value representing text or alternatively a second binary value representing background for the text; and 2. means for applying a graphics threshold to graphics blocks in the scalar image file to convert each scaled pixel graphic value into a first binary value representing one color or alternatively a second binary value representing another color wherein the graphics threshold application means includes: (a) means for passing a dither matrix over the graphics blocks to re-assign a background pixel value to all scalar pixel values in the block less than a corresponding dither value in the dither matrix; (b) means for passing an error diffuse dither matrix over the graphics blocks to yield error diffuse pixel values; (c) comparing the graphics threshold value to the difference between the non-re-assigned scalar pixel values and the corresponding error diffuse pixel values; and (d) means for re-setting the scalar pixel value to a first binary number if the difference is greater than the graphics threshold and for otherwise re-setting the scalar pixel value to a second binary number; and D. means for compressing the binary image file to yield a compressed image file for subsequent decompression and digital processing of the compressed image file.