US20030107770A1 — Algorithm for adjusting edges of grayscale pixel-…

Claim 1

Independent Claim 267 words

1 . A sizing algorithm for sizing a parent grayscale pixel map expressing an image having edges using a computer, before projection onto a sensitive recording surface wherein the pixels have a size R p , comprising the steps of: (i) inputting the parent grayscale pixel map image with edges where an edge is defined by gray pixels having values between 1, 2, . . . n, or by pixels having at least one 0-gray value (black 0-dose level) pixel neighbor; (ii) calculating a grayscale correction value (g) equal to a sizing distance S parameterized by a machine constant equal to R p divided by the number of grayscale values; (iii) finding and flagging edge pixels expressed within a frame of the parent pixel map; and (iv) finding and flagging corner edge pixels within the frame; (v) sliding a sub-matrix window within the frame, to calculate and store gradient values for each edge pixel relative to the edges within the frame; (vi) looping over pixels within the frame to adjust the grayscale value of edge and corner pixels and neighboring pixels; (vii) propagating new grayscale values per the grayscale correction value (g) to pixels from each adjusted edge and comer pixel within the frame in a direction normal to each edge to establish a new edge position within the frame, and (viii) where the parent pixel map is composed of a plurality of frames, reassembling the frames generating a daughter grayscale pixel map expressing a different size image than that expressed in the parent pixel map system which upon projection and recording, compensates for expected systemic distortions.

Claim 2

Dependent Claim 267 words

1 . A sizing algorithm for sizing a parent grayscale pixel map expressing an image having edges using a computer, before projection onto a sensitive recording surface wherein the pixels have a size R p , comprising the steps of: (i) inputting the parent grayscale pixel map image with edges where an edge is defined by gray pixels having values between 1, 2, . . . n, or by pixels having at least one 0-gray value (black 0-dose level) pixel neighbor; (ii) calculating a grayscale correction value (g) equal to a sizing distance S parameterized by a machine constant equal to R p divided by the number of grayscale values; (iii) finding and flagging edge pixels expressed within a frame of the parent pixel map; and (iv) finding and flagging corner edge pixels within the frame; (v) sliding a sub-matrix window within the frame, to calculate and store gradient values for each edge pixel relative to the edges within the frame; (vi) looping over pixels within the frame to adjust the grayscale value of edge and corner pixels and neighboring pixels; (vii) propagating new grayscale values per the grayscale correction value (g) to pixels from each adjusted edge and comer pixel within the frame in a direction normal to each edge to establish a new edge position within the frame, and (viii) where the parent pixel map is composed of a plurality of frames, reassembling the frames generating a daughter grayscale pixel map expressing a different size image than that expressed in the parent pixel map system which upon projection and recording, compensates for expected systemic distortions.

Claim 3

Dependent Claim 22 words

2 . The algorithm of claim 1 where the frame is a 5×5 matrix, and the sub-matrix window is a 3×3 matrix.

Claim 4

Dependent Claim 162 words

3 . A sizing algorithm for downsizing a parent grayscale pixel map having pixels of size R p and grayscale values between 0, 1, 2, . . . n, expressing an image having edges with grayscale values ranging from 1, 2, . . . n, using a computer, before projection onto a sensitive recording surface comprising the steps of: calculating a factor g equal to a desired sizing distance S divided by a machine constant K m equal to pixel size R p divided by the number of grayscale values; setting G′(i, j) = Max(G(i, j) − g,0), and {right arrow over (δ)}G(i, j) = |G(i, j) − g|·(∇ x , ∇ y ) looping over pixels { if( pixel ij is an edge pixel) { ∇(i, j) = estimated gradient; storing the value G′(i, j) of the new pixel; if(∥{right arrow over (δ)}G(i, j)∥) > 0 { propagating vector differences to neighboring pixels along the gradient direction; } } } end.

Claim 5

Dependent Claim 162 words

3 . A sizing algorithm for downsizing a parent grayscale pixel map having pixels of size R p and grayscale values between 0, 1, 2, . . . n, expressing an image having edges with grayscale values ranging from 1, 2, . . . n, using a computer, before projection onto a sensitive recording surface comprising the steps of: calculating a factor g equal to a desired sizing distance S divided by a machine constant K m equal to pixel size R p divided by the number of grayscale values; setting G′(i, j) = Max(G(i, j) − g,0), and {right arrow over (δ)}G(i, j) = |G(i, j) − g|·(∇ x , ∇ y ) looping over pixels { if( pixel ij is an edge pixel) { ∇(i, j) = estimated gradient; storing the value G′(i, j) of the new pixel; if(∥{right arrow over (δ)}G(i, j)∥) > 0 { propagating vector differences to neighboring pixels along the gradient direction; } } } end.

Claim 6

Dependent Claim 164 words

4 . A sizing algorithm for upsizing a parent grayscale pixel map having pixels of size R p and grayscale values between 0, 1, 2, . . . n, expressing an image having edges with grayscale values ranging from 1, 2, . . . n, using a computer, before projection onto a sensitive recording surface comprising the steps of: calculating a factor g equal to a desired sizing distance S divided by a machine constant K m equal to pixel size R p divided by the number of grayscale values; setting G′(i, j) = Min(G(i, j) + g,gmax), and {right arrow over (δ)}G(i, j) = |gmax − {G(i, j) + g}|·(∇ x , ∇ y ) looping over pixels { if( pixel ij is an edge pixel) { ∇(i, j) = estimated gradient; storing the value G′(i, j) of the new pixel; if(∥{right arrow over (δ)}G(i, j)∥) > 0 { propagating vector differences to neighboring pixels along the gradient direction; } } } end.

Claim 7

Dependent Claim 164 words

4 . A sizing algorithm for upsizing a parent grayscale pixel map having pixels of size R p and grayscale values between 0, 1, 2, . . . n, expressing an image having edges with grayscale values ranging from 1, 2, . . . n, using a computer, before projection onto a sensitive recording surface comprising the steps of: calculating a factor g equal to a desired sizing distance S divided by a machine constant K m equal to pixel size R p divided by the number of grayscale values; setting G′(i, j) = Min(G(i, j) + g,gmax), and {right arrow over (δ)}G(i, j) = |gmax − {G(i, j) + g}|·(∇ x , ∇ y ) looping over pixels { if( pixel ij is an edge pixel) { ∇(i, j) = estimated gradient; storing the value G′(i, j) of the new pixel; if(∥{right arrow over (δ)}G(i, j)∥) > 0 { propagating vector differences to neighboring pixels along the gradient direction; } } } end.

Claim 8

Dependent Claim 103 words

5 . The sizing algorithm of claim 1 , or 3 or 4 wherein edge pixels are flagged by successively mapping a sub matrix array G of pixel grayscale values from the parent grayscale pixel map into a an edge matrix E with a Boolean procedure for counting the 0-gray value pixels and for assigning a value to each pixel of 0, 1, or 2, where 0 indicates a particular pixel is not an edge pixel, 1 indicates a particular pixel is in a class consisting of inclined edge pixels and corner pixels, and 2 indicates a particular pixel is an edge pixel.

Claim 9

Dependent Claim 18 words

6 . The sizing algorithm of claim 5 wherein the sub matrix array G is a 3×3 matrix.

Claim 10

Dependent Claim 50 words

7 . The sizing algorithm of claim 6 wherein diverging corner edge pixels are flagged by successively mapping each edge matrix E into a Boolean I returning true if E has an edge_sum equal to 5 indicating a particular edge pixel is a diverging corner edge pixel, else returning false.

Claim 11

Dependent Claim 99 words

8 . The sizing algorithm of claim 5 wherein new grayscale values are propagated normal to edges of the parent grayscale pixel map to new edge positions with: (i) a 45° rule computation operator for such edges inclined at 45° (π/4) relative to an orthogonal coordinate of the parent pixel map; and (ii) a non-45 20 rule computation operator for such edges inclined at angles other than 45° relative to an orthogonal coordinate of the parent pixel map; and (iii) a gray-to-neighbors computation operator for propagating gray values to pixels neighboring such edges in the direction of the gradient.

Algorithm for adjusting edges of grayscale pixel-map images

Abstract

Inventors

Bibliographic Details

Claims Analysis

References Cited (Backward Citations)

USPTO Annuity & Maintenance Windows

Assigned Categories & Tags

Assign Category Tag