Abstract Due to an ever-increasing population, crowd density estimation is an extremely significant factor that must be considered in todays times, and people counting is a critical subject in video surveillance applications. When a large group of people gathers, there is a risk of loss of life, property, and other things. Human action identification, crowd anomaly detection, and behavior analysis are some of the most popular disciplines of video processing study. This is where crowd density estimate, which is based on image and video processing, can help organizers of public events, railway security, college campuses, and other places maintain track of crowd density. Various image processing and video processing techniques are employed in this project effort to estimate the number of persons from a given video footage or image. Keywords: Crowd Detection, deep learning, image processing, video processing