##go.tesseract go.tesseract is a wrapper for the tesseract OCR library (text-recognition from image/pdf).
go.tesseract has two direct dependencies; go.leptonica and libtesseract
Make sure you have installed go.leptonica. go.leptonica has a C library dependency, please read the go.leptonica/README.md.
You are required to install the tesseract library including development headers at version 3.02.02 or later. You absolutely need 3.02.02 (or later) as go.tesseract can not compile with earlier versions of tesseract. At time of writing this version of tesseract is not in the ubuntu/debian stable repository yet.
go.tesseract uses gopkg.in for versioned releases:
go get gopkg.in/GeertJohan/go.tesseract.v1
sudo apt-get install -t testing libtesseract3 libtesseract-dev
Do the following before trying to go get this package:
$ brew install leptonica
$ brew install tesseract
$ export CGO_LDFLAGS="-L$(brew --prefix leptonica)/lib -L$(brew --prefix tesseract)/lib"
$ export CGO_CFLAGS="-I$(brew --prefix leptonica)/include -I$(brew --prefix tesseract)/include"
Note: this assumes you are using the standard Brew path of /usr/local/Cellar
Download, configure, make and install
svn checkout http://tesseract-ocr.googlecode.com/svn/tags/release-3.02.02 tesseract-ocr-read-only
cd tesseract-ocr-read-only
./autogen.sh
./configure
make
sudo make install
sudo ldconfig
If you have installed from debian testing (jessie):
sudo apt-get install -t testing tesseract-ocr-YOUR-LANGUAGE-SHORTCODE
# example, this installs dutch and english
sudo apt-get install -t testing tesseract-ocr-nld
sudo apt-get install -t testing tesseract-ocr-eng
If you have installed manually; copy language files (do this for any language you require)
sudo cp tessdata/YOUR-LANGUAGE-SHORTCODE.* /usr/local/share/tessdata/
# example for english and dutch:
sudo cp tessdata/eng.* /usr/local/share/tessdata/
sudo cp tessdata/nld.* /usr/local/share/tessdata/
For more information, view the tesseract compilation guide.