A Surveillance with a Geographic Information System to count crowd in real-time using a Deep Convolution Neural Network with Drone Technology

DOI

10.2991/978-94-6463-662-8_14How to use a DOI?

Keywords

Crowd Density; Geographic Information System (GIS); Deep Learning; Convolutional Neural Network (CNN); Drone Technology; Real-time Surveillance; Smart City

Abstract

Advanced urbanization processes and the growing scale of public events suggest the need for fast and accurate crowd monitoring and density assessment tools to provide safety, manage resources, and respond to emergency situations. This work presents an advanced approach that utilizes GIS with DCNN and drones to present a real-time solution for crowd monitoring and surveillance. Using the feature extraction ability of CNNs, the system achieves correct density map generation from aerial imagery of drones capturing noisy density maps, thus enhancing the reliability of crowd monitoring even in cases with occlusion and varying illumination. This is supported by the GIS platform which provides map-based analysis and visualization tools, for real-time decision-making and interventions. All the deep learning frameworks are tested with four architectures namely, CNN, InceptionResNetV2, MobileNet, and highest performing EfficientNetB0 to determine the architecture suitable for real-time applications. The experimental outcomes show that CNN models yield lower MAE and MSE values than the other models, and MobileNet and EfficientNetB0 can be considered as solution-efficient lightweight models. The integration of drones guarantees more coverage and effective movement in spatial terms making the system much flexible with high mobility in various sectors like smart city, disaster response and management, and event surveillance. Furthermore, real-time GIS-based mapping and Image overlay enable the integration of aerial data whereby stakeholders are assisted in identifying areas with density and possible risk areas. Three important issues that the proposed system would solve include data fusion, variability in the environment as well as resource limitation, making the proposed system a portable, flexible, intelligent system that would fit the needs of contemporary crowds management.

Copyright

© 2025 The Author(s)

Open Access

Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

Download article (PDF)

TY  - CONF
AU  - S. Vinay Kumar
AU  - V. Suresh
AU  - O. Sirisha
AU  - G. K. Nagaraju
PY  - 2025
DA  - 2025/03/17
TI  - A Surveillance with a Geographic Information System to count crowd in real-time using a Deep Convolution Neural Network with Drone Technology
BT  - Proceedings of the International Conference on Advanced Materials, Manufacturing and Sustainable Development (ICAMMSD 2024)
PB  - Atlantis Press
SP  - 162
EP  - 181
SN  - 2352-5401
UR  - https://doi.org/10.2991/978-94-6463-662-8_14
DO  - 10.2991/978-94-6463-662-8_14
ID  - Kumar2025
ER  -