Compile tesseract for iOS SDK 6.0

http://lois.di-qual.net/blog/compile-tesseract-for-ios-sdk-6-0/点击打开链接

Update 10/25/12

Added instructions to integrate an armv7s slice in liblept.a and libtesseract_all.a.

Update 09/24/12

To make the iOS integration easier, I made a Github repo containing an Objective-C wrapper for Tesseract. It relies on an other repo containing tesseract & leptonica libs. Feel free to improve them !

• Objective-C wrapper: https://github.com/ldiqual/tesseract-ios
• Compiled libraries: https://github.com/ldiqual/tesseract-ios-lib

More informations about these repos on my next article.

Introduction

I had troubles while following the compiling guide from this blog:http://tinsuke.wordpress.com/2011/11/01/how-to-compile-and-use-tesseract-3-01-on-ios-sdk-5/. Some dependencies are missing on a fresh Mountain Lion with Xcode 4.5, and more importantly: the build_dependencies.sh won’t work.

Here is why: Xcode 4.5 and iOS SDK 6.0 don’t support armv6 anymore.

So here are the steps to build tesseract-ocr for iOS SDK 6.0.

Requirements

If you don’t have HomeBrew already, that’s a shame. Run to this website and install it.

Then, install libtool and automake:

1	$ brew install libtool automake

You should have installed ‘Command Line Tools’ after the Homebrew installation, but if you forgot: Open Xcode’s preferences, and install them from the ‘Downloads’ tab.

The build process will need libtoolize. Just make a symlink to glibtoolize:

1	$ ln -s /usr/local/bin/glibtoolize /usr/local/bin/libtoolize

We’re good to go.

Sources and compilation

The steps will more or less follow the tinsuke’s article, except for one thing: the build script.

Get the following sources:

• tesseract: http://code.google.com/p/tesseract-ocr/downloads/detail?name=tesseract-3.01.tar.gz
• leptonica: http://www.leptonica.com/source/leptonica-1.69.tar.gz
• build_dependencies.sh: https://gist.github.com/3763217 (click on “raw” and save it to your computer)

Then, organize your folder with the following structure:

1 2 3 4 5

./build/   dependencies/    leptonica-1.69/   tesseract-3.01/   build_dependencies.sh

You might want to modify build_dependencies.sh to make LEPTON_LIB and TESSERACT_LIB match the proper folders.

Now, you can build the libraries:

1 2	$ chmod +x ./build_dependencies.sh $ ./build_dependencies

In dependencies/lib/, you’ll find liblept.a and libtesseract_all.a, these are the only libraries you need so you can remove other ones. Both contain armv7 and i386 slices only, but many iOS 5 projects require an armv7s slice.

Just follow this Matt Galloway’s blog post to generate and put an armv7s slice into the libs.

Now you just have to copy dependencies/ in your Xcode Project, and follow my next article on how to use tesseract.