AboutIt is convenient to use your smartphone to take picture of papers of documents, and then the texts in the picture can be recognized and translated automatically within seconds. This project consists of a client application on Android and cloud services, and provides cloud-based recognition and translation of text image on Android devices.
Known IssueOnly English texts can be recognized. You can choose other language or "Auto Dectect" in the "Setting" in Menu. There is, however, no any effect. By default the texts will be translated into Japanese. You can choose other languages in the "Setting" in Menu. Illumination will greatly impact the recognized result. Even a small shadow in the picture can cause the failure of recognition. Complex background will cause the faulure of recognition. The best result is achieved if the background is white and text is black. This project is still in beta phase. Please report bugs in the issue tab. Any suggestions or comments are welcomed. More detailsThe recognition engine is powered by OCRopus, and the translation is by Google Translate. The client application take shot of the text in the documents, send the image to cloud server, get result from the server, and show result on your screen. The servers cosists of both frontend server and backend server. The frontend is powered by Google App Engine, and the backend server is actually Python script which calls OCRopus to perform recognition.
Use Patent Claims
These details are provided for information only. No information here is legal advice and should not be used as such.