Modern life is tech-savvy. Today’s man cannot live without using modern gadgets. Among them smartphone has become an easy accessory to share the emotions, attitudes, pleasures and pains of the man by constantly taking videos that attract the soul of the man. Hence a study on the techniques of video summarisation is indispensable. Video summarisation has become a compelling field of research due to quick growth in video content in a number of sectors, such as social media, education, entertainment, and surveillance. With an emphasis on both abstractive and extractive tactics, the present study examines different approaches and methodologies developed for video summarising. Key frames or segments from the original video are identified during the extractive summarisation process using techniques like shot boundary recognition and clustering. Conversely, abstractive summarisation uses machine learning models such as “generative adversarial networks, multi-modal learning, reinforcement learning, attention mechanism, deep neural networks, and natural language processing” to excerpt the most important information from video and produce new content. In this paper it also examines the applications in practical implementations. The datasets used to benchmark these methods are also covered in the study. The objective of study is to bring out an up-to-date, comprehensive understanding of case and potential future paths of video summarisation research.