VIDEO TEXT DETECTION AND CONVERSION TO AUDIO FILE

VIDEO TEXT DETECTION AND CONVERSION TO AUDIO FILE

SUDEEP GOWDA S R 1, VIDYA S 2

1 Student, Dept. of MCA, Bangalore Institute of Technology, Karnataka, India

2 Assistant Professor, Dept. of MCA, Bangalore Institute of Technology, Karnataka, India

---------------------------------------------------------------------***---------------------------------------------------------------------

Abstract -Video text detection and recognition play a crucial role in extracting meaningful information from multimedia content for various applications such as video indexing, content retrieval, and automatic captioning. This paper provides a comprehensive study on the application of a R-CNN and CRNN algorithm for video text detection and recognition. In an era dominated by video content, the fusion of Optical Character Recognition (OCR), faster R-CNN for Text Detection and Convolutional Recurrent Neural Network (CRNN) for text recognition holds the promise of transforming the way we interact with and extract insights from visual media. This project aims to provide a comprehensive system for video text detection, OCR implementation, R-CNN and CNN meaningful text output generation. The core objective is to bridge the gap of visual information and textual insights, enabling automated extraction, analysis, and classification of text within video frames. By seamlessly integrating advanced OCR techniques, the system extracts textual content from video frames, overcoming challenges posed by diverse fonts, sizes, and backgrounds. The extracted text is then subjected to pre-processing to ensure accuracy and relevance.

Key Words:  Binirizationan, frame-extraction, Grey scale conversion, Noise reduction Dilation, Convolutional Recurrent Neural Network (CRNN), Faster R-CNN