1Pnmtojbig User Manual(0) Pnmtojbig User Manual(0)
2
3
4
6 pnmtojbig - PNM to JBIG file converter
7
8
10 pnmtojbig
11
12 [options] [input-file [output-file]]
13
14
16 This program is part of Netpbm(1).
17
18 pnmtojbig reads a PBM or PGM image, compresses it, and outputs the
19 image as a JBIG bi-level image entity (BIE) file.
20
21 JBIG is a highly effective lossless compression algorithm for bi-level
22 images (one bit per pixel), which is particularly suitable for scanned
23 document pages.
24
25 A JBIG encoded image can be stored in several resolutions (progressive
26 mode). These resolution layers can be stored all in one single BIE or
27 they can be stored in several separate BIE files. All resolution lay‐
28 ers except the lowest one are stored merely as differences to the next
29 lower resolution layer, because this requires less space than encoding
30 the full image completely every time. Each resolution layer has twice
31 the number of horizontal and vertical pixels than the next lower layer.
32 JBIG files can also store several bits per pixel as separate bitmap
33 planes, and pnmtojbig can read a PGM file and transform it into a
34 multi-bitplane BIE.
35
36
38 -q Encode the image in one single resolution layer (sequential
39 mode). This is usually the most efficient compression method. By
40 default, the number of resolution layers is chosen automatically
41 such that the lowest layer image is not larger than 640 × 480
42 pixels.
43
44
45 -x number
46 Specify the maximal horizontal size of the lowest resolution
47 layer. The default is 640 pixels.
48
49
50 -y number
51 Specify the maximal vertical size of the lowest resolution
52 layer. The default is 480 pixels.
53
54
55 -l number
56 Select the lowest resolution layer that will be written to the
57 BIE. It is possible to store the various resolution layers of a
58 JBIG image in progressive mode into different BIEs. Options -l
59 and -h allow you to select the resolution-layer interval that
60 will appear in the created BIE. The lowest resolution layer has
61 number 0 and this is also the default value. By default, pnmto‐
62 jbig writes all layers.
63
64
65 -h number
66 Select the highest resolution layer that will be written to the
67 BIE. By default, pnmtojbig writes all layers. See also option
68 -l.
69
70
71 -b Use binary values instead of Gray code words in order to encode
72 pixel values in multiple bitplanes. This option has only an
73 effect if the input is a PGM file and if more than one bitplane
74 is produced. Note that the decoder has to make the same selec‐
75 tion but cannot determine from the BIE, whether Gray or binary
76 code words were used by the encoder.
77
78
79 -d number
80 Specify the total number of differential resolution layers into
81 which the input image will be split in addition to the lowest
82 layer. Each additional layer reduces the size of layer 0 by 50
83 %. This option overrides options -x and -y, which are usually a
84 more comfortable way of selecting the number of resolution lay‐
85 ers.
86
87
88 -s number
89 The JBIG algorithm splits each image into a number of horizontal
90 stripes. This option specifies that each stripe shall have num‐
91 ber lines in layer 0. The default value is selected so that
92 approximately 35 stripes will be used for the whole image.
93
94
95 -m number
96 Select the maximum horizontal offset of the adaptive template
97 pixel. The JBIG encoder uses a number of neighbour pixels in
98 order to get statistical a priori knowledge of the probability,
99 whether the next pixel will be black or white. One single pixel
100 out of this template of context neighbor pixels can be moved
101 around. Especially for dithered images it can be a significant
102 advantage to have one neighbor pixel which has a distance large
103 enough to cover the period of a dither function. By default,
104 the adaptive template pixel can be moved up to 8 pixels away.
105 This encoder supports up to 23 pixels, however as decoders are
106 only required to support at least a distance of 16 pixels by the
107 standard, no higher value than 16 for number is recommended in
108 order to maintain interoperability with other JBIG implementa‐
109 tions. The maximal vertical offset of the adaptive template
110 pixel is always zero.
111
112
113 -t number
114 Encode only the specified number of most significant bit planes.
115 This option allows to reduce the depth of an input PGM file if
116 not all bits per pixel are needed in the output.
117
118
119 -o number
120 JBIG separates an image into several horizontal stripes, resolu‐
121 tion layers and planes, were each plane contains one bit per
122 pixel. One single stripe in one plane and layer is encoded as a
123 data unit called stripe data entity (SDE) inside the BIE. There
124 are 12 different possible orders in which the SDEs can be stored
125 inside the BIE and number selects which one shall be used. The
126 order of the SDEs is only relevant for applications that want to
127 decode a JBIG file which has not yet completely arrived from
128 e.g. a slow network connection. For instance some applications
129 prefer that the outermost of the three loops (stripes, layers,
130 planes) is over all layers so that all data of the lowest reso‐
131 lution layer is transmitted first.
132
133 The following values for number select these loop arrangements
134 for writing the SDEs (outermost loop first):
135
136
137
138 0 planes, layers, stripes
139
140
141 2 layers, planes, stripes
142
143
144 3 layers, stripes, planes
145
146
147 4 stripes, planes, layers
148
149
150 5 planes, stripes, layers
151
152
153 6 stripes, layers, planes
154
155
156
157 All loops count starting with zero, however by adding 8 to the
158 above order code, the layer loop can be reversed so that it
159 counts down to zero and then higher resolution layers will be
160 stored before lower layers. Default order is 3 which writes at
161 first all planes of the first stripe and then completes layer 0
162 before continuing with the next layer and so on.
163
164
165 -p number
166 This option allows you to activate or deactivate various
167 optional algorithms defined in the JBIG standard. Just add the
168 numbers of the following options which you want to activate in
169 order to get the number value:
170
171
172
173
174 4 deterministic prediction (DPON)
175
176
177 8 typical prediction (TPBON)
178
179
180 16 diff. layer typical prediction (TPDON)
181
182
183 64 layer 0 two-line template (LRLTWO)
184
185
186
187 Except for special applications (like communication with JBIG
188 subset implementations) and for debugging purposes you will nor‐
189 mally not want to change anything here. The default is 28,
190 which provides the best compression result.
191
192
193 -c The adaptive template pixel movement is determined as suggested
194 in annex C of the standard. By default the template change
195 takes place directly in the next line which is most effective.
196 However a few conformance test examples in the standard require
197 the adaptive template change to be delayed until the first line
198 of the next stripe. This option selects this special behavior,
199 which is normally not required except in order to pass some con‐
200 formance test suite.
201
202
203 -v After pnmtojbig creates the BIE, it lists a few technical
204 details of the created file (verbose mode).
205
206
207
208
210 Most of the format pnmtojbig creates is defined by the JBIG standard.
211
212 The standard, however, does not specify which values in the BIE mean
213 white and which mean black. It contains a recommendation that for a
214 single plane image zero mean background and one mean foreground, but
215 the Netpbm formats have no concept of foreground and background. And
216 the standard says nothing about values for multiple plane BIEs.
217
218 pnmtojbig follows Markus Kuhn's implementation of the standard in the
219 pbmtojbg program that comes with his JBIG library: If the BIE is a
220 single plane BIE, zero means white and one means black. If it is a
221 multiple plane BIE, zero means black and the maximal value is white.
222
223
225 This program implements the JBIG image coding algorithm as specified in
226 ISO/IEC 11544:1993 and ITU-T T.82(1993).
227
228
230 pnmtojbig is based on the JBIG library by Markus Kuhn, part of his
231 JBIG-KIT package. The pbmtojbg program is part of the JBIG-KIT pack‐
232 age. The most recent version of that library and tools set is freely
233 available on the Internet from anonymous ftp server ftp.informatik.uni-
234 erlangen.de ⟨ftp://ftp.informatik.uni-erlangen.de⟩ in directory
235 pub/doc/ISO/JBIG/.
236
237 pnmtojbig is part of the Netpbm package of graphics tools.
238
239
241 pnm(1), jbigtopnm(1)
242
243
245 If you use pnmtojbig, you are using various patents, particularly on
246 its arithmetic encoding method, and in all probability you do not have
247 a license from the patent owners to do so.
248
249
250
251netpbm documentation 20 May 2000 Pnmtojbig User Manual(0)