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 "Pnmsmooth User Manual" 0 "19 December 2009" "netpbm documentation"
7
8.SH NAME
9
10pnmsmooth - smooth out an image
11
12.UN synopsis
13.SH SYNOPSIS
14
15\fBpnmsmooth\fP
16[\fB-width=\fP\fIcols\fP] [\fB-height=\fP\fIrows\fP]
17[\fIpnmfile\fP] [\fB-size\fP]
18.PP
19Minimum unique abbreviations of options is acceptable.  You may use double
20hyphens instead of single hyphen to denote options.  You may use white
21space in place of the equals sign to separate an option name from its value.
22
23.UN description
24.SH DESCRIPTION
25.PP
26This program is part of
27.BR "Netpbm" (1)\c
28\&.
29.PP
30\fBpnmsmooth\fP smoothes out an image by replacing each pixel with the
31average of its width X height neighbors.  It is implemented as a progam that
32invokes \fBpnmconvol\fP with an appropriate convolution matrix.
33
34.UN options
35.SH OPTIONS
36
37
38.TP
39\fB-width=\fP\fIcols\fP
40.TP
41\fB-height=\fP\fIrows\fP
42These options specify the dimensions of the convolution matrix.
43Default dimensions are 3 wide and 3 high.
44.sp
45Before Netpbm 10.49 (December 2009), the maximum size of the convolution
46matrix is limited by the maxval of the image such that width * height * 2 must
47not exceed the maxval.  (use \fBpamdepth\fP to increase the maxval if
48necessary).
49.sp
50These options were new in Netpbm 10.32 (February 2006).  Before that,
51use \fB-size\fP.
52
53.TP
54\fB-size\fP
55This deprecated option exists in current Netpbm for backward
56compatibility.  It was obsoleted by \fB-width\fP and \fB-height\fP
57in Netpbm 10.32 (February 2006).
58.sp
59When you use this option, the first two program arguments are the width
60and height, respectively, of the convolution matrix and do the same thing
61as the \fB-width\fP and \fB-height\fP option values.  The third
62(optional) program argument is the input file name.
63.sp
64In reality, in old \fBpnmsmooth\fP, the width and height are two
65values of the \fB-size\fP option, but the modern Netpbm command syntax
66paradigm doesn't allow an option with multiple values, so instead
67\fB-size\fP is an option with no value and width and height are program
68arguments.  That has the fortunate effect of making the following command
69mean the same in current \fBpnmsmooth\fP as in old \fBpnmsmooth\fP:
70.nf
71\f(CW
72     pnmsmooth -size 5 5 infile.ppm >outfile.ppm
73\fP
74.fi
75
76.TP
77\fB-dump=\fP\fIdumpfile\fP
78This options makes \fBpnmsmooth\fP only show you the
79convolution matrix.  It writes to Standard Output a \fBpnmconvol\fP
80\fB-matrix\fP option value that represents the matrix.  It does not
81invoke \fBpnmconvol\fP and does not produce an output image.
82.sp
83Before Netpbm 10.49 (December 2009), this option is rather different.
84It takes a file name as a value, and it writes to that file the
85convolution matrix as a PGM file (as used to be the normal input for
86\fBpnmconvol\fP).
87
88
89
90
91
92.UN seealso
93.SH SEE ALSO
94.BR "pnmconvol" (1)\c
95\&,
96.BR "pnm" (5)\c
97\&
98
99.UN history
100.SH HISTORY
101.PP
102Before Netpbm 10.32 (February 2006), \fBpnmsmooth\fP did not use
103the modern Netpbm command line parser, so had an unconventional command line
104syntax.  Most importantly, you could not use an equal sign or double
105hyphens in the options.
106.PP
107Before Netpbm 10.49 (December 2009), there was a \fB-dump\fP option.
108This strange option caused \fBpnmsmooth\fP not to do any smoothing or
109produce any output image but instead write the convolution matrix it
110would have used, as PGM file such as \fBpnmconvol\fP used to use, to
111a file you specify.  The idea was you could then use that file with a
112separate invocation of \fBpnmconvol\fP.
113.PP
114Then, in Netpbm 10.49, there was a rather different \fB-dump\fP
115option with a similar purpose: It caused \fBpnmsmooth\fP to write to
116Standard Error a string suitable as a value for the \fBpnmconvol\fP
117\fB-matrix\fP option (an option that was new in Netpbm 10.49).
118.PP
119But in Netpbm 10.51 (June 2010), \fBpnmconvol\fP started using the even
120newer \fBpnmconvol\fP \fB-normalize\fP option (new in 10.50), which made
121specifying the convolution matrix for the kind of smoothing that
122\fBpnmsmooth\fP does trivial, so \fB-dump\fP disappeared from
123\fBpnmsmooth\fP.
124.PP
125(There were also ease of implementation issues that kept us from simply
126keeping the original \fB-dump\fP around for backward compatibility: As we
127modified \fBpnmsmooth\fP to take advantage of the new features of
128\fBpnmconvol\fP, which \fBpnmsmooth\fP uses internally, the information
129needed to implement \fB-dump\fP was no longer available in the program).
130.SH DOCUMENT SOURCE
131This manual page was generated by the Netpbm tool 'makeman' from HTML
132source.  The master documentation is at
133.IP
134.B http://netpbm.sourceforge.net/doc/pnmsmooth.html
135.PP