Cricket Video Summarization Using Deep Learning

In this paper, we present a novel approach to summarizing cricket match videos using deep learning techniques. Leveraging computer vision and deep learning algorithms, our methodology focuses on extracting key events such as sixes, fours, and wickets from cricket match footage. The process involves multiple stages, including frame extraction, scorecard detection, text extraction, bowler identification, and event boundary determination. We employ state-of-the-art deep learning models, such as YOLOv8, for object detection and Optical Character Recognition (OCR) for text extraction. By combining these techniques, we generate a comprehensive summary of the match, highlighting significant moments and key plays. Our approach not only automates the summarization process but also enhances the accessibility and comprehensibility of cricket match highlights for viewers. By conducting thorough experimentation and juxtaposing our approach with established methodologies, we showcase the efficacy and expediency of our innovative framework in crafting compelling and informative video synopses.