A Framework for Character Recognition APP Using ML Kit

Abstract

Character recognition applications are pivotal in enabling real-time translation and digitization of printed and handwritten text. Text recognition can be implemented using a variety of technologies found in the field of software development, but here for Android mobile development, the ML Kit is used. The goal is to create an Android character recognition app, especially for Devanagari script (with language converter software included as a feature), using the ML kit without Firebase that will make recognizing, learning, and language translation easier and will promote stress-free communication. Using ML Kit is a boon for developers without extensive knowledge of machine learning, as it simplifies the integration of complex ML features and saves significant time in learning and implementation. This app has the ability to recognise in real time as well as storage-based recognition on both handwritten and printed scripts. The app incorporates real-time recognition and offline translation features, supporting recognition in six languages and translating text into 59 languages. This app will perform more effectively than other existing programmes since it will use optimized code for the translation and recognition process. The major goal is to have the software operate even when it is not connected to the internet.