We tried the Google Docs API for a long time, but it is very weak in terms of accuracy. It seems to be based on some outdated version of tesseract. I guess you get more accuracy if you try tesseract. However, you will need to control the special preprocessing of images taken by the camera, as they create additional problems. The Google Docs API does not have this.
When launching Tesseract on Android, see here: Using tesseract for android
The primary tesseract alternative for OCR on a mobile phone: http://www.abbyy.com/mobileocr/
However, if you are looking for not only text capture, but also data extraction, you may need additional technology to analyze text output. This means writing even more code. Or there is an alternative to licensing the existing Commertial Data Capture API from ABBYY. This has already been discussed here:
Recognize text at a specific position using the iPhone camera
Disclaimer: I work for ABBYY
source share