Human Crowd Counting

Abstract

Human detection nowadays has become a necessary part of our lives from a security point of view as well as for our ease. The more increasing the world population, the higher the sense of detection of the crowd will be required. In this paper, we have combined all the techniques and methodologies proposed by different review papers. Several research papers have introduced diverse methods for addressing the challenge of pedestrian counting. These approaches encompass feature-based, trajectory-based, map-based, and hardware-based techniques. Additionally, certain studies employ colour image processing to accurately count pedestrians, while others explore the utilization of stereo cameras to monitor individuals as they pass. Some innovative approaches involve the detection of humans using various gases, and various algorithms for counting humans have been explored in the literature. It is noteworthy that the human ability to count verbally in discrete quantities is a distinct cognitive capability when compared to animal cognition, adding to the complexity of these counting methodologies. Moreover, some research endeavours leverage Wi-Fi technology for human detection, while others incorporate the use of three distinct cameras to enhance the accuracy of human detection processes. There are several approaches and strategies used to determine the human count. Human count sounds like a simple task, but it is more complex than it seems. It is mainly done for crowd safety and to create routes for humans or crowds to exit different areas in case of emergency. It can also be used to predict the size of future gatherings. From a business point of view, analysing or estimating the expected crowd identifies a further analysis of demand and supply and thus allows business owners to be more equipped. Human detection is mostly used in public places and high-security areas such as malls, temples, hotels, clubs, institutions, airports, railway stations, etc.