Optical Character Recognition App

The OCR feature is super-cool and surprisingly good -- it actually was able to recognize most of the characters in a (neatly u6977u4e66) calligraphed reprint of one of the early u7ea2u697cu68a6 manuscripts -- but if you've got enough Chinese to be familiar with radicals and stroke order or if you've got Apple's basic handwriting input system and don't mind spending the one second it takes to draw stuff in yourself then it's really not crucial. Which Pleco add-ons are or aren't crucial sort of depends on where you are in your study of Chinese. If you're able to use Chinese-Chinese dictionaries the u6c49u8bedu89c4u833u8bcdu5178 is worth the price of purchase; if you prefer Chinese-English dictionaries but need something with more oomph than the built-in dictionary the ABC Chinese-English dictionary is great for general use though it's missing a lot of newer vocabulary. Supposedly Pleco will be releasing a new pack of dictionaries at some point this year and I'm looking forward to that; ABC and the Guifan dictionary are the only real must-haves from the current set of options I think. OCR is super-neat but I don't think I've ever actually used it in anger -- though then again I'm a translator rather than an active student of Chinese so my use cases may be different from yours. I will say that the handwriting input add-on for Pleco which licenses handwriting recognition tech from Hanwang is awesome and totally worth the price of admission if handwritten input is something you do with any regularity. It's absolutely miles better than Apple's though supposedly iOS 6 will feature improvements to Chinese input and is more forgiving of cursive input nonstandard stroke order and all the real-world issues that Apple's system tends to fall down on.

Hi I am answering your question assuming the app that you are intending to make is not just restricted to a particular mobile device. You could use Python - use pytesseract package. C++ - use Asprise C++ SDK or tesseract ocr Java - Asprise has Java bindings too. Python C++ and Java have their advantages. Decide for yourself what best suits your purpose! Hope this was helpful.

Personally i used scanner pro best app. I tried marathi English and hindi ocr works well but sometimes it doesn't work properly. Ps - check the attachment

It is very difficult to detect the with tessearct. You can use Stroke Width Transform to detect the from the image . Then use blob or contour to extract the region. Apply tessearct with hindi language mode. In tessearct you will get confidence score for each word. If the input word is hindi you will get High confidence score. You can use the confidence score to classify whether the input word is hindi or English.