Transcription of Text
Pramila P. Shinde, Niyoshi Soni, Dhruwangi Shrotriya, Sahil Shivhare · Zenodo (CERN European Organization for Nuclear Research) · 2022
Optical Character Recognition (OCR) is a technology that recognizes text in documents and converts it into an editable machine-readable format. Feature extraction, recognition, and classification into appropriate labels are the primary components of an OCR system. This paper uses image processing and (OCR) based architecture to segment, recognize, and identify documents. Moreover, handwriting has evolved, as evidenced by the different types of handwritten characters such as digit, numeral, cursive script, and symbols English and other languages. The automatic recognition of text can be an extremely useful application in healthcare, education and personal care. The construction of text recognition systems is possible with the help of Easy OCR which has some basic components of data acquisition, processing and Query and visualization. As the name implies, Easy OCR is a Python package that allows computer vision professionals to do Optical Character Recognition with ease. Additionally, our model “Text Transcription” also translates the detected text into user desirable language and generates an audio output to ease out the reading from user’s end by making it more functional and user-friendly. It is two-way system which also does the conversion of speech to text.