Description:
OSRA is a utility designed to convert graphical representations of chemical structures, as they appear in journal articles, patent documents, textbooks, trade magazines etc., into SMILES (Simplified Molecular Input Line Entry Specification - see http://en.wikipedia.org/wiki/SMILES) or SD file - a computer recognizable molecular structure format. OSRA can read a document in any of the over 90 graphical formats parseable by ImageMagick - including GIF, JPEG, PNG, TIFF, PDF, PS etc., and generate the SMILES or SDF representation of the molecular structure images encountered within that document.
Note that any software designed for optical recognition is unlikely to be perfect, and the output produced might, and probably will, contain errors, so a curation by a human knowledgeable in chemical structures is highly recommended.

News:

• Starting with version 1.2.1 there is a windows installer which automatically installs a plugin if Symyx Draw is present. It also detects and auto installs if necessary Ghostscript and GflAx libraries.
• OSRA manuscript has been published: "Optical Structure Recognition Software To Recover Chemical Information: OSRA, An Open Source Solution" J. Chem. Inf. Model., 2009, 49 (3), pp 740–743.
• Starting with version 1.2.0 plugins for BKChem, MolSketch, Symyx Draw, and Scitegic PipelinePilot are now included with Windows zip archive. Plugins allow for integration of OSRA functionality with chemical structure editors and other chemoinformatics software.

• ImageMagick, image manipulation library, version 6.2.8 or later; if installing from RPM make sure you have the following packages:
  ImageMagick
  ImageMagick-devel
  ImageMagick-c++-devel
  ImageMagick-c++
  http://www.imagemagick.com/
• POTRACE, vector tracing library, version 1.7 or later,
  http://potrace.sourceforge.net/
• GOCR/JOCR, optical character recognition library, version 0.43 or later - NOTE version 0.46 is NOT RECOMMENDED! It lacks the necessary libPgm2asc library,
  http://jocr.sourceforge.net/
• OCRAD, optical character recognition program, version 0.17,
  http://www.gnu.org/software/ocrad/ocrad.html
• TCLAP, Templatized C++ Command Line Parser Library, version 1.1.0,
  http://tclap.sourceforge.net/
• OpenBabel, open source chemistry toolbox, version 2.2.0 or later; if installing from RPM make sure you have the following packages:
  openbabel
  openbabel-devel
  http://openbabel.sourceforge.net/wiki/Main_Page
• OR alternatively you can use RDKit as the molecular builder back-end, version Q32008_1 or later (svn snapshot is recommended at the moment):
  http://www.rdkit.org/

• ThinImage, C code from the article "Efficient Binary Image Thinning using Neighborhood Maps" by Joseph M. Cychosz, 3ksnn64@ecn.purdue.edu in "Graphics Gems IV", Academic Press, 1994
  http://www.acm.org/pubs/tog/GraphicsGems/gemsiv/thin_image.c
• GREYCstoration, Anisotropic smoothing plugin,
  http://www.greyc.ensicaen.fr/~dtschump/greycstoration/
• CImg, The C++ Template Image Processing Library,
  http://cimg.sourceforge.net
• MCDL utility from Sergei Trepalin and Andrei Gakh for 2D coordinate generation

• Computer-generated 2D structures, such as found on the PubChem website, http://pubchem.ncbi.nlm.nih.gov/, black-and-white and color (use a resolution of 72 dpi),
• Black-and-white PDF and PostScript files, including multi-page ones. Please note that you need Ghostcript installed for ImageMagick to be able to parse these kinds of files. OSRA internally renders PS and PDF at a resolution of 150 dpi, so you do not need to specify the resolution as a commmand-line parameter,
• Scanned images - black-and-white, a resolution of 300 dpi is recommended, though 150 dpi can also produce fair results. Please make sure the scanned image is of reasonable quality - an input that's too noisy will only generate garbage output.

Command-line options:
./osra --help
will give you a list of available options with short descriptions.

Most common use: ./osra [-r <resolution>] <filename>
Resolution in dpi, default is 300 (unless it's a PS or PDF file as mentioned above), filename is the name of your image file (or PS/PDF document).

Other options:
-t, --threshold: Gray level threshold, default is 0.2 for black-and-white images,
-n, --negate: Inverts colors (for white on black images),
-o, --output: Sets a prefix for writing recognized images to files - i.e. "-o tmp" will create files tmp0.png, tmp1.png... for each of the structures,
-s, --size: Resize images on output - can be useful for running OSRA as a backend for a webservice. Example: "-s 300x400".
-g, --guess: Prints out resolution guess when you chose to have automatic resolution estimate.
-p, --print: Prints out the value of confidence function estimate.
-f, --format: Output format (either smi for SMILES or sdf for SD file format)

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., 59 Temple Place, Suite 330, Boston, MA 02111-1307, USA. See also http://www.gnu.org/.

See the file COPYING for details.