Debian Codegraph @

upstream/debian-unstable/man1/pgmkernel.1

raw
1 \
2 .\" This man page was generated by the Netpbm tool 'makeman' from HTML source.
3 .\" Do not hand-hack it! If you have bug fixes or improvements, please find
4 .\" the corresponding HTML page on the Netpbm website, generate a patch
5 .\" against that, and send it to the Netpbm maintainer.
6 .TH "Pgmkernel User Manual" 1 "19 December 2013" "netpbm documentation"
7
8 .SH NAME
9 pgmkernel - generate a convolution kernel
10
11 .UN synopsis
12 .SH SYNOPSIS
13 .PP
14 \fBpgmkernel\fP
15 [\fB-weight=\fP\fInumber\fP]
16 [\fB-maxval=\fP] {\fIsize\fP | \fIwidth\fP \fIheight\fP}
17 .PP
18 Minimum unique abbreviation of option is acceptable. You may use double
19 hyphens instead of single hyphen to denote options. You may use white
20 space in place of the equals sign to separate an option name from its value.
21
22
23 .UN description
24 .SH DESCRIPTION
25 .PP
26 This program is part of
27 .BR "Netpbm" (1)\c
28 \&.
29 .PP
30 \fBpgmkernel\fP generates a convolution kernel that you can use
31 with \fBpnmconvol\fP. The kernel is one where the weight of each location
32 is inversely proportional to its distance from the center of the kernel.
33 .PP
34 \fBpgmkernel\fP generates a PGM image of width and height \fIsize\fP
35 if you specify one argument, or width \fIwidth\fP and height \fIheight\fP
36 if you specify two arguments.
37 .PP
38 \fBpgmkernel\fP computes the convolution function K as follows.
39
40 .RS
41 K(i,j) = 1 / ( 1 + w * sqrt(i^2 + j^2))
42 .RE
43
44 where \fIw\fP is a coefficient specified via the \fB-weight\fP
45 option. \fIi\fP and \fIj\fP are measured in pixels. K is zero
46 everywhere beyond the specified kernel width and height.
47 .PP
48 The sample values in the PGM output have this value scaled and biased using
49 the protocol \fBpnmconvol\fP specifies for representing the real numbers K
50 in PGM.
51
52
53 .UN options
54 .SH OPTIONS
55 .PP
56 In addition to the options common to all programs based on libnetpbm
57 (most notably \fB-quiet\fP, see
58 .UR index.html#commonoptions
59 Common Options
60 .UE
61 \&), \fBpgmkernel\fP recognizes the following
62 command line options:
63
64
65 .TP
66 \fB-weight=\fP\fInumber\fP
67 The distance from the center is weighted by this; a higher number means
68 the value falls off more quickly as you go away from the center.
69 .sp
70 This must be a positive real number.
71 .sp
72 The default is 6.0.
73
74 .TP
75 \fB-maxval=\fP\fImaxval\fP
76 The maxval for the PGM kernel.
77 .sp
78 Default is 255.
79 .sp
80 This option was new in Netpbm 10.65 (December 2013). Before that, the
81 maxval is always 255.
82
83
84
85
86 .UN limitations
87 .SH LIMITATIONS
88 .PP
89 The computation time is proportional to \fIwidth\fP*\fIheight\fP.
90 This increases rapidly with the increase of the kernel size. A better
91 approach could be using a FFT in these cases.
92
93 .UN history
94 .SH HISTORY
95 .PP
96 Before Netpbm 10.65 (December 2013), the output was always in
97 Plain (text) PGM format. (Now, like standard Netpbm programs, the default
98 is raw PGM and you can get Plain PGM with a \fB-plain\fP option).
99 .PP
100 Before Netpbm 10.65 (December 2013), this manual said negative values
101 for \fB-weight\fP were valid (as long as they were greater than -1.0). But
102 the program never worked with negative numbers and it isn't clear that the
103 result would be useful, so \fB-weight\fP is now required to be nonnegative
104 and the program fails gracefully if you specify a negative value.
105
106 .UN seealso
107 .SH SEE ALSO
108 .BR "pnmconvol" (1)\c
109 \&,
110 .BR "pnmsmooth" (1)\c
111 \&
112 .BR "pamgauss" (1)\c
113 \&
114 .BR "pgm" (1)\c
115 \&
116
117
118 .UN author
119 .SH AUTHOR
120
121 Alberto Accomazzi (\[email protected]\fP).
122 .SH DOCUMENT SOURCE
123 This manual page was generated by the Netpbm tool 'makeman' from HTML
124 source. The master documentation is at
125 .IP
126 .B http://netpbm.sourceforge.net/doc/pgmkernel.html
127 .PP