Character Recognition and Extraction using OpenCV and Pytesseract

Abstract—This paper introduces an optimized framework for real-time Optical Character Recognition (OCR) and video-to- text conversion using OpenCV and Tesseract, by implementing adaptive preprocessing techniques and efficient frame handling strategies, the system achieves high text detection accuracy while significantly reducing processing time. Additionally, a multi- threaded audio transcription process enhances performance for converting video content into text. The proposed approach is well- suited for applications requiring real-time text extraction from multimedia sources, such as automated document processing and video surveillance.. Index Terms—OCR, Pytesseract, Cv2