Aedict3 OCR is a stand-alone application which performs OCR (optical character recognition) of Japanese text, preferably on a simple white background. You can store multiple pages of Japanese text, gradually perform OCR over parts of these pages, correct the OCRed text and analyze the text with Aedict3.
Aedict3 OCR uses an off-line OCR engine. Such engines uses a lot of processor power, which in turn will drain your battery faster and may heat up your phone in extreme cases. The algorithm is good, but it is not perfect and it may mis-find some characters. You will get best results for a black-on-white sharp images, with uniform white background, such as printouts. The algorithm is probably next to unusable on color-y magazine covers and hand-drawn kanjis (found for example on temples).
Aedict3 OCR works on Android 4.0 and greater.
- Aedict3 OCR remembers OCRed images and OCRed text for multiple pages
- Text auto-correction and manual correction (ば -> ぱ, つっ etc)
- Back-up and restore of images and texts, to the internal phone memory, ownCloud and Dropbox
Read/Write External storage: used when importing images from Gallery
Network: used when backing up/restoring data to/from ownCloud and Dropbox