A People Counting System Using Convolutional Neural Network for Avoiding Public Gathering and Maintaining Social Distancing

Image processing and computer vision are two essential components of object identification. It is utilized in a variety of technological and economic domains, including laboratory, agriculture, research, and manufacturing, as well as medical diagnosis. In this work, we discussed the various strategies as well as the issues encountered in picture categorization and object counting. It is a difficult technique that is dependent on a variety of circumstances. Our goal is to investigate various approaches utilized in perception and item counting. Furthermore, we addressed specific use cases of this module in real-world applications. One of the biggest challenges in the retail industry is measuring how many customers enter a store throughout the day. For example in a pandemic situation likely, that occurred in the form of covid-19 few years back, maintaining track of people movements and social distancing rules in a commercial setting is crucial in the current circumstances because only way to stay safe in this situation is social distancing. In order to lessen the possibility of disease transmission, social distancing is a public health technique that seeks to keep sick people from being in close proximity to healthy people. It is a system where we use a webcam or we can give our video or images to count the number of people in a frame and if the number of people in the frame exceeds a certain limit, then it gives an alert. In this work, we are using for human detection. It is an object detection algorithm which is trained on the image dataset. The detections are very fast and thus can be used in a real-time scenario.