Provided by: potrace_1.13-2_amd64 bug


       potrace - transform bitmaps into vector graphics.


       potrace [options] [filename...]


       potrace  is a tool for tracing a bitmap, which means, transforming a bitmap into a smooth,
       scalable image.  The input is a bitmap, which means, a pixel-based image composed  of  the
       two  colors black and white only. The output is EPS, PDF, SVG, or one of a number of other
       vector formats. A typical use is to create vector graphics  from  scanned  data,  such  as
       company  or  university  logos, handwritten notes, etc. The resulting image is not "jaggy"
       like a bitmap, but smooth. It can then be rendered at any resolution.

       potrace can read bitmaps in the following formats: PBM, PGM, PPM  (collectively  known  as
       PNM, see pnm(5)), as well as BMP (Windows and OS/2 bitmap formats). The input image should
       only use the two colors black and white. If other pixel values appear in the  input,  they
       will be converted to black and white using a simple threshold method.

       potrace  can  currently  produce  the following output formats: EPS, PostScript, PDF, SVG,
       DXF, GeoJSON, PGM, Gimppath, and XFig.  Additional backends might be added in the future.


       The following options are supported. Dimensions (arguments of type dim) can have  optional
       units,  e.g.  6.5in, 15cm, 100pt.  The default unit is inches (or centimeters, if this was
       configured at compile time, see COMPILE TIME CONFIGURATION below). For pixel-based  output
       formats such as PGM, DXF, GeoJSON, and Gimppath, the default unit is pixels.

   General options:
       -h, --help     print help message and exit.

       -v, --version  print  version  info  and  exit.  This  also  shows  the defaults that were
                      compiled into this version of potrace.

       -l, --license  print license info and exit.

   Input/output options:
       filename       Each file can hold an input image, or multiple concatenated  input  images.
                      If  filename  arguments  are given, then potrace will by default create one
                      output file for each input filename given. The name of the output  file  is
                      obtained  from  the  input filename by changing its suffix according to the
                      chosen backend. If changing the suffix is impossible because the  names  of
                      the  input and output files would be identical, then the output filename is
                      created by adding the "-out" suffix to the name of the input  file.  If  no
                      filename  arguments  are given, then potrace acts as a filter, reading from
                      standard input and writing to standard output. A filename  of  "-"  may  be
                      given to specify reading from standard input.

       -o filename, --output filename
                      write output to this file. All output is directed to the specified file. If
                      this option is used, then multiple input filenames  are  only  allowed  for
                      multi-page  backends  (see  BACKEND  TYPES below). In this case, each input
                      file may contain one or more bitmaps, and all  the  bitmaps  from  all  the
                      input files are processed and the output concatenated into a single file. A
                      filename of "-" may be given to specify writing to standard output.

       --             End of options. Any remaining arguments are interpreted as filenames.  This
                      also  disables  filter mode, even if no filenames are given. This is useful
                      for shell scripts, because potrace -- $FILENAMES will behave correctly even
                      for an empty list of filenames. However, -- with an empty list of filenames
                      is not permitted in conjunction with the  -o  option,  because  this  would
                      generate a document of zero pages, which none of the backends permit.

   Backend selection:
       For general information, see also BACKEND TYPES below.

       -b name, --backend name
                      Select  backend  by  name,  where  name is one of eps, postscript, ps, pdf,
                      pdfpage, svg, dxf, geojson, pgm,  gimppath,  xfig.  Backend  names  can  be
                      abbreviated  by  a  prefix  as long as it is unambiguous. Backend names are
                      case insensitive.

       -e, --eps, -b eps, --backend eps
                      EPS backend (default). The output is an encapsulated PostScript file.  This
                      is a single-page, variable-sized, dimension-based backend.

       -p, --postscript, -b ps, --backend ps
                      PostScript  backend. The output is a PostScript file. This is a multi-page,
                      fixed-size, dimension-based backend. If  the  input  consists  of  multiple
                      bitmaps, they are each rendered on a separate page.

       -b pdf, --backend pdf
                      PDF  backend. The output is a file in the Portable Document Format.  If the
                      input consists of multiple bitmaps, they are each rendered  on  a  separate
                      page. This is a multi-page, variable-sized, dimension-based backend.

       -b pdfpage, --backend pdfpage
                      The  PDFPage  backend is like the PDF backend, except that it is fixed-size
                      like the PostScript backend.

       -s, --svg, -b svg, --backend svg
                      SVG backend. The output is a Scalable Vector Graphics (SVG) file.  This  is
                      a  single-page,  variable-sized,  dimension-based backend. Note that unless
                      the -r option is given, the resolution of the input bitmap is assumed to be

       -b dxf, --backend dxf
                      DXF  backend. The output is a file in the Drawing Interchange Format (DXF).
                      In this backend, all Bezier curves are approximated by  piecewise  circular
                      arcs;  this  is  suitable  for  processing in CAD software or for machining
                      applications using CNC tools. This is a single-page, variable-sized, pixel-
                      based backend. The -u option has no effect for this backend.

       -b geojson, --backend geojson
                      GeoJSON  backend.  The  output  is  a  file  in  the  format  used  by some
                      applications processing geographical data.  In  this  backend,  all  Bezier
                      curves are approximated by 8 straight line segments. This is a single-page,
                      variable-sized, pixel-based backend. The -u option has no effect  for  this

       -g, --pgm, -b pgm, --backend pgm
                      PGM  backend.  The  output  is  a  portable  greymap  (PGM)  file.  It is a
                      convenient backend for antialiasing a bitmap image. This is  a  multi-page,
                      variable-sized, pixel-based backend. If the input consists of more than one
                      image, the images are concatenated in the output.

       -b gimppath, --backend gimppath
                      Gimppath backend. This backend produces output suitable to be imported as a
                      path  by the GNU Image Manipulation Program (Gimp) (in the Layers, Channels
                      & Paths dialog, select Paths, then right-click and select Import Path). The
                      output is actually an SVG file. The differences to the SVG backend are: the
                      --opaque option has no effect, the --flat option  is  always  on,  and  the
                      dimensions  are  pixel-based. This is a single-page, variable-sized, pixel-
                      based backend.

       -b xfig, --backend xfig
                      XFig backend. This is a single-page, fixed-size,  dimension-based  backend.
                      The  output  is  a  file in the XFig format.  Note that XFig uses X-splines
                      instead of Bezier curves, thus it is not possible to translate  the  output
                      of potrace into the XFig format with absolute accuracy. This backend does a
                      reasonable approximation using two control points  for  each  Bezier  curve
                      segment.  The  -u  option  has  no effect for this backend, because control
                      points are always rounded to the nearest 1/1200 of an inch in  XFig.  Curve
                      optimization is disabled. Implies --opaque.

   Algorithm options:
       For more detailed information on these options, see TECHNICAL DOCUMENTATION below.

       -z policy, --turnpolicy policy
                      specify  how  to  resolve ambiguities in path decomposition. Must be one of
                      black, white, right,  left,  minority,  majority,  or  random.  Default  is
                      minority.  Turn policies can be abbreviated by an unambiguous prefix, e.g.,
                      one can specify min instead of minority.

       -t n, --turdsize n
                      suppress speckles of up to this many pixels.

       -a n, --alphamax n
                      set the corner threshold parameter. The default value  is  1.  The  smaller
                      this  value,  the more sharp corners will be produced. If this parameter is
                      0, then no smoothing will be performed and the output is a polygon. If this
                      parameter  is  greater  than  4/3,  then all corners are suppressed and the
                      output is completely smooth.

       -n, --longcurve
                      turn off curve optimization. Normally potrace tries to join adjacent Bezier
                      curve  segments  when this is possible. This option disables this behavior,
                      resulting in a larger file size.

       -O n, --opttolerance n
                      set the curve optimization tolerance. The  default  value  is  0.2.  Larger
                      values  allow  more consecutive Bezier curve segments to be joined together
                      in a single segment, at the expense of accuracy.

       -u n, --unit n set output quantization. Coordinates in the output are  rounded  to  1/unit
                      pixels. The default of 10 usually gives good results. For some of the debug
                      modes, a value of 100 gives more accurate output. This option has no effect
                      for  the  XFig  backend, which always rasterizes to 1/1200 inch, or for the
                      DXF backend. For the GeoJSON backend, this  option  is  only  a  hint;  the
                      actual rounding may be more, but not less, accurate than specified.

       -d n, --debug n
                      produce  debugging  output  of  type  n.  This  has  different  effects for
                      different backends. For the PostScript/EPS  backends,  the  values  n=1,2,3
                      illustrate the intermediate stages of the potrace algorithm.

   Scaling and placement options:
       -P format, --pagesize format
                      for  fixed-size  backends,  set  page  size.  The  following formats can be
                      specified: A4, A3, A5, B5, Letter, Legal,  Tabloid,  Statement,  Executive,
                      Folio,  Quarto, 10x14. Format names are case insensitive. Also, an argument
                      of the form dimxdim  is  accepted  to  specify  arbitrary  dimensions.  The
                      default page size is Letter (or A4, if this was configured at compile time,
                      see  COMPILE  TIME  CONFIGURATION  below).   Page  format  names   can   be
                      abbreviated  by  a  prefix as long as it is unambiguous. This option has no
                      effect for variable-sized backends.

       -W dim, --width dim
                      set the width of output image (before any rotation and  margins).  If  only
                      one  of width and height is specified, the other is adjusted accordingly so
                      that the aspect ratio is preserved.

       -H dim, --height dim
                      set the height of output image. See -W for details.

       -r n[xn], --resolution n[xn]
                      for dimension-based backends, set the resolution (in dpi). One inch in  the
                      output  image  corresponds  to  this  many pixels in the input. Note that a
                      larger value results in a smaller output image.  It is possible to  specify
                      separate resolutions in the x and y directions by giving an argument of the
                      form nxn. For variable-sized backends, the default resolution is 72dpi. For
                      fixed-size  backends,  there  is  no  default  resolution;  the image is by
                      default scaled to fit on the page. This option has  no  effect  for  pixel-
                      based backends. If -W or -H are specified, they take precedence.

       -x n[xn], --scale n[xn]
                      for  pixel-based  backends,  set the scaling factor. A value greater than 1
                      enlarges the output, a value between 0 and 1 makes the output smaller.  The
                      default  is 1. It is possible to specify separate scaling factors for the x
                      and y directions by giving an argument of the form nxn. This option has  no
                      effect  for  dimension-based backends. If -W or -H are specified, they take

       -S n, --stretch n
                      set the aspect ratio. A value greater  than  1  means  the  image  will  be
                      stretched  in the y direction. A value between 0 and 1 means the image will
                      be compressed in the y direction.

       -A angle, --rotate angle
                      set  the  rotation  angle  (in  degrees).  The  output  will   be   rotated
                      counterclockwise  by this angle. This is useful for compensating for images
                      that were scanned not quite upright.

       -M dim, --margin dim
                      set all four margins. The effect and default value of this option depend on
                      the backend.  For variable-sized backends, the margins will simply be added
                      around the output image (or subtracted, in case of negative  margins).  The
                      default  margin  for  these  backends  is  0.  For fixed-size backends, the
                      margin settings can be used to control the placement of the  image  on  the
                      page.  If only one of the left and right margin is given, the image will be
                      placed this distance from the respective edge of the  page,  and  similarly
                      for  top  and  bottom. If margins are given on opposite sides, the image is
                      scaled to  fit  between  these  margins,  unless  the  scaling  is  already
                      determined  explicitly  by one or more of the -W, -H, -r, or -x options. By
                      default, fixed-size backends use a non-zero margin whose width  depends  on
                      the page size.

       -L dim, --leftmargin dim
                      set the left margin. See -M for details.

       -R dim, --rightmargin dim
                      set the right margin. See -M for details.

       -T dim, --topmargin dim
                      set the top margin. See -M for details.

       -B dim, --bottommargin dim
                      set the bottom margin. See -M for details.

       --tight        remove  whitespace around the image before scaling and margins are applied.
                      If this option is given, calculations of the width, height, and margins are
                      based  on the actual vector outline, rather than on the outer dimensions of
                      the input pixmap, which is the default. In particular, the  --tight  option
                      can  be  used  to remove any existing margins from the input image. See the
                      file placement.pdf for a more detailed illustration.

   Color options:
       These options are only supported by certain backends. The DXF and GeoJSON backends do  not
       support color.

       -C #rrggbb, --color #rrggbb
                      set the foreground color of the output image. The default is black.

       --fillcolor #rrggbb
                      set  the  fill  color  of  the output image, i.e., the color of the "white"
                      parts that are enclosed by "black" parts. The default  is  to  leave  these
                      parts transparent. Implies --opaque.  Please note that this option sets the
                      background color; to set the foreground color, use --color instead.

       --opaque       fill in the white parts of the image  opaquely,  instead  of  leaving  them
                      transparent.  This  only  applies to interior white parts, i.e., those that
                      are enclosed inside a black outline. Opaqueness is always in effect for the
                      XFig backend.

   SVG options:
       --group        for  SVG  output, try to group related paths together. Each path is grouped
                      together with all paths that are contained inside it, so that they  can  be
                      moved  around  as a unit with an SVG editor. This makes coloring individual
                      components slightly more cumbersome, and thus it is not the default.

       --flat         for SVG output, put the entire image into a  single  path.  This  makes  it
                      impossible  to  color  the  components individually, and thus it is not the
                      default. But the resulting SVG file can be more  easily  imported  by  some
                      applications  such as Gimp. In fact, the Gimppath backend is a variation of
                      the SVG backend with the --flat option and pixel-based scaling. The  --flat
                      option has no effect if --opaque has been selected.

   PostScript/EPS/PDF options:
       -c, --cleartext
                      do  not  compress  the  output. This option disables the use of compression
                      filters in the PostScript and PDF output. In the PostScript backend, if  -c
                      and  -q are used together, the resulting output can be easily read by other
                      programs or even by humans.

       -2, --level2   use PostScript level 2 compression (default). The resulting  file  size  is
                      ca. 40% smaller than if the -c option is used.

       -3, --level3   use  PostScript  level  3  compression,  if  available. This gives slightly
                      smaller files than using -2, but the resulting files may not print on older
                      PostScript  level 2 printers. If support for PostScript level 3 compression
                      has been disabled at compile time, a warning message is printed and level 2
                      compression is used instead.

       -q, --longcoding
                      turn off optimized numerical coding in PostScript output. Normally, potrace
                      uses a  very  compact  numerical  format  to  represent  Bezier  curves  in
                      PostScript,   taking   advantage   of  existing  redundancy  in  the  curve
                      parameters. This option disables this behavior, resulting  in  longer,  but
                      more readable output (particularly if the -c option is also used).

   PGM options:
       -G n, --gamma n
                      set  the  gamma  value  for  anti-aliasing  (default is 2.2). Most computer
                      displays do not render shades of grey linearly, i.e., a grey value  of  0.5
                      is  not  displayed  as  being exactly half-way between black and white. The
                      gamma parameter corrects for this, and therefore  leads  to  nicer  looking
                      output.  The  default  value  of  2.2  is  appropriate  for most normal CRT

   Frontend options:
       -k n, --blacklevel n
                      set the threshold level for converting input images to bitmaps. The potrace
                      algorithm  expects  a  bitmap,  thus  all  pixels  of  the input images are
                      converted to  black  or  white  before  processing  begins.   Pixels  whose
                      brightness  is  less  than  n  are  converted to black, all other pixels to
                      white. Here n is a number between 0 and 1. One case is  treated  specially:
                      if  the input is in an indexed color format with exactly 2 colors, then the
                      blacklevel is ignored and the darker of the two colors is mapped to black.

                      Note: the method used by potrace for converting greymaps to bitmaps is very
                      crude;  much  better results can be obtained if a separate program, such as
                      mkbitmap(1), is used for this purpose. In particular, mkbitmap(1), which is
                      distributed  with  potrace,  has  the  ability to scale and interpolate the
                      image before thresholding, which results in  much  better  preservation  of

       -i, --invert   invert the input bitmap before processing.

   Progress bar options:
       --progress     display  a  progress  bar for each bitmap that is processed. This is useful
                      for interactive use.  The default behavior is  not  to  show  any  progress

       --tty mode     set  the  terminal  mode  for  progress  bar rendering. Possible values are
                      "vt100", which requires a vt100-compatible terminal, and "dumb", which uses
                      only ASCII characters. The default is system dependent.


       Backends  can  be  classified  in  several  ways, which affects the available command line
       options and their behavior:

       Fixed-size or variable-sized:
            For fixed-size backends, the size of the page is always the same (for example  Letter
            or  A4, as specified at compile time or by the -P option). By default, the image will
            be centered and scaled to fit the page size. For variable-size backends, the size  of
            the  page  follows the size of the image. Currently the PostScript (PS), PDFPage, and
            XFig backends are fixed-size, and the remaining backends are variable-size.

       Dimension-based or pixel-based:
            In dimension-based backends, distances are measured in physical units such as  inches
            or  centimeters.  In pixel-based backends, distances are measured in pixel units. The
            -r option only works for dimension-based backends, and the -x option only  works  for
            pixel-based  backends.  Currently,  the  DXF, PGM, Gimppath, and GeoJSON backends are
            pixel-based, and the remaining backends are dimension-based.  Currently,  all  pixel-
            based backends are variable-sized.

       Single-page or multi-page:
            Single-page  backends  can only accept a single image. Multi-page backends can accept
            multiple images, typically one per page of output. Currently,  the  PostScript  (PS),
            PDF, PDFPage, and PGM backends are multi-page, and the remaining backends are single-
            page. Note that multiple input images can be read in two ways:  from  multiple  input
            files  (with  the  -o  option),  or  from  a  single  input  file  that holds several
            concatenated images.


       Certain aspects of the behavior of potrace can be configured at compile  time  by  passing
       the following options to the ./configure script.

            compile  potrace  without the zlib compression library. This means PostScript level 3
            compression will not be available.

            compile potrace with centimeters as the default unit instead of inches.

            compile potrace with A4 as the default page size.


       The exit status is 0 on successful completion, 1 if the command line was invalid, and 2 on
       any other error.




       Peter Selinger <selinger at>

       Please see the file AUTHORS for a full list of other contributors.


       For  a  detailed technical description of the potrace algorithm, see the file potrace.pdf,
       which is available from the potrace web site. For information on the Potrace library  API,
       see potracelib.pdf.


       The latest version of potrace is available from This site
       also contains a list of frequently asked questions, as  well  as  information  on  how  to
       obtain support.




       Copyright (C) 2001-2015 Peter Selinger

       This program is free software; you can redistribute it and/or modify it under the terms of
       the GNU General Public License as  published  by  the  Free  Software  Foundation;  either
       version 2 of the License, or (at your option) any later version.

       This  program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY;
       without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR  PURPOSE.
       See the GNU General Public License for more details.

       You should have received a copy of the GNU General Public License along with this program;
       if not, write to the Free Software Foundation, Inc.,  51  Franklin  Street,  Fifth  Floor,
       Boston, MA 02110-1301, USA.  See also