1Pamditherbw User Manual(0) Pamditherbw User Manual(0)
2
3
4
6 pamditherbw - dither grayscale image to black and white
7
8
10 pamditherbw
11
12 [-floyd | -fs | -atkinson | -threshold | -hilbert | -dither8 | -d8 |
13 -cluster3 | -c3 | -cluster4 | -c4 | -cluster8 | -c8]
14
15 [-value val]
16
17 [-clump size]
18
19 [-randomseed=integer]
20
21 [pamfile]
22
23 All options can be abbreviated to their shortest unique prefix.
24
25
27 This program is part of Netpbm(1).
28
29 pamditherbw dithers a grayscale image. Dithering means turning each
30 shade of gray into a pattern of black and white pixels that, from a
31 distance, look the same as the gray.
32
33 The input should be a PGM image or a PAM image of tuple type GRAYSCALE.
34 However, pamditherbw doesn't check, so if you feed it e.g. a PPM image,
35 it will produce arbitrary results (actually, it just takes the first
36 channel of whatever you give it and treats it as if it represented gray
37 levels).
38
39 The output is a PAM with tuple type BLACKANDWHITE. You can turn this
40 into a PBM (if you need to use it with an older program that doesn't
41 understand PAM) with pamtopnm.
42
43 To do the opposite of dithering, you can usually just scale the image
44 down and then back up again with pamscale, possibly smoothing or blur‐
45 ring the result with pnmsmooth or pnmconvol. Or use the special case
46 program pbmtopgm.
47
48 To dither a color image (to reduce the number of pixel colors), use
49 ppmdither.
50
51 Another way to convert a grayscale image to a black and white image is
52 thresholding. Thresholding is simply replacing each grayscale pixel
53 with a black or white pixel depending on whether its brightness is
54 above or below a threshold. That threshold might vary. Simple thresh‐
55 olding is a degenerate case of dithering, so pamditherbw does very sim‐
56 ple thresholding with its -threshold option. But pamthreshold does
57 more sophisticated thresholding.
58
59 If all you want is to change a PGM image with maxval 1 to a PBM image,
60 pamtopnm will do that.
61
62
63
65 The default quantization method is boustrophedonic Floyd-Steinberg
66 error diffusion (-floyd or -fs).
67
68 Also available are simple thresholding (-threshold); Bayer's ordered
69 dither (-dither8) with a 16x16 matrix;
70 Atkinson
71 ⟨http://www.tinrocket.com/projects/programming/graphics/00158/⟩ ; and
72 three different sizes of 45-degree clustered-dot dither (-cluster3,
73 -cluster4, -cluster8).
74
75 A space filling curve halftoning method using the Hilbert curve is also
76 available (-hilbert).
77
78 Floyd-Steinberg or Atkinson will almost always give the best looking
79 results; however, looking good is not always what you want. For
80 instance, you can use thresholding in a pipeline with the pnmconvol,
81 for tasks such as edge and peak detection. And clustered-dot dithering
82 gives a newspaper-ish look, a useful special effect.
83
84 Floyd-Steinberg is by far the more traditional, but
85 some claim
86 ⟨http://www.tinrocket.com/projects/programming/graphics/00158/⟩ Atkin‐
87 son works better.
88
89 The Hilbert curve method is useful for processing images before display
90 on devices that do not render individual pixels distinctly (like laser
91 printers). This dithering method can give better results than the
92 dithering usually done by the laser printers themselves. The -clump
93 option alters the number of pixels in a clump. Typically a PGM image
94 will have to be scaled to fit on a laser printer page (2400 x 3000 pix‐
95 els for an A4 300 dpi page), and then dithered to a PBM image before
96 being converted to a postscript file. A printing pipeline might look
97 something like:
98
99 pamscale -xysize 2400 3000 image.pgm | pamditherbw -hilbert | \
100 pamtopnm | pnmtops -scale 0.25 > image.ps
101
102
103
104
105 -value This option alters the thresholding value for Floyd-Steinberg,
106 Atkinson, and simple thresholding. It should be a real number
107 between 0 and 1. Above 0.5 means darker images; below 0.5 means
108 lighter.
109
110
111 -clump This option alters the number of pixels in a clump. This is
112 usually an integer between 2 and 100 (default 5). Smaller clump
113 sizes smear the image less and are less grainy, but seem to lose
114 some grey scale linearity.
115
116
117 -randomseed=integer
118 This is the seed for the random number generator that generates
119 the pixels.
120
121 Use this to ensure you get the same image on separate invoca‐
122 tions.
123
124 By default, pgmnoise uses a seed derived from the time of day
125 and process ID, which gives you fairly uncorrelated results in
126 multiple invocations.
127
128 This option was new in Netpbm 10.45 (December 2008).
129
130
131
132
133
135 The only reference you need for this stuff is "Digital Halftoning" by
136 Robert Ulichney, MIT Press, ISBN 0-262-21009-6.
137
138 The Hilbert curve space filling method is taken from "Digital Halfton‐
139 ing with Space Filling Curves" by Luiz Velho, Computer Graphics Volume
140 25, Number 4, proceedings of SIGRAPH '91, page 81. ISBN 0-89791-436-8
141
142
144 pamtopnm(1), pgmtopgm(1), pbmtopgm(1), pamthreshold(1), pbmreduce(1),
145 pnmconvol(1), pamscale(1), pam(1), pnm(1),
146
147
149 pamditherbw was new in Netpbm 10.23 (July 2004), but is essentially the
150 same program as pgmtopbm that has existed practically since the begin‐
151 ning. pamditherbw differs from its predecessor in that it properly
152 adds brightnesses (using gamma transformations; pgmtopbm just adds them
153 linearly) and that it accepts PAM input in addition to PGM and PBM and
154 produces PAM output.
155
156 pamditherbw obsoletes pgmtopbm.
157
158 -atkinson was new in Netpbm 10.38 (March 2007).
159
160
162 Copyright (C) 1989 by Jef Poskanzer.
163
165 This manual page was generated by the Netpbm tool 'makeman' from HTML
166 source. The master documentation is at
167
168 http://netpbm.sourceforge.net/doc/pamditherbw.html
169
170netpbm documentation 10 May 2010 Pamditherbw User Manual(0)