Downloads: 6 | Views: 42 | Weekly Hits: ⮙2 | Monthly Hits: ⮙3
Research Paper | Computer Science & Engineering | India | Volume 13 Issue 6, June 2024 | Rating: 5.7 / 10
Multi Column Convolutional Neural Network for Accurate Crowd Counting and Analysis in Highly Congested Urban Scenes
Abstract: The estimation of highly congested, highly varied crowded scenes is a challenging vision task that has received a lot of interest in recent years. Crowd counting and analysis aims to count the number of people and make an analysis of the density of the crowded scene. Exponential growth in the world population and the resulting urbanization has led to an increase in the number of activities such as sporting events, political rallies, public demonstrations which would thereby result in a more frequent crowd gathering. In such situations, it is essential to analyze crowd behavior for better management, safety and security. In this paper, a Convolutional Neural Network (CNN) based approach is used for the application. Among the various approaches, Multi Column Convolutional Neural Network is used to train the network in order to estimate the number of people. Also, the crowd analysis task is associated with many challenges such as non-uniform density, intra-scene and inter-scene variations in scale, occlusions and perspective.
Keywords: Convolutional Neural Network, Crowd counting, Digital image processing
Edition: Volume 13 Issue 6, June 2024,
Pages: 724 - 730