Real-Time Army Surveillance System Using IOT and Deep Learning Techniques

Border security is an important task for every country, as borders are always sensitive to different types of threats. In many areas, surveillance is still done using manual patrolling and basic camera systems. These methods need continuous human attention and sometimes fail to detect suspicious activities on time, especially in remote border locations. Because of these issues, there is a need for a better and more reliable surveillance system. This project presents an army surveillance system using IoT and deep learning for real-time monitoring of border areas. A PC-based webcam is used to capture live video from the border region. The captured video frames are processed using the YOLOv11 algorithm to detect people present in the monitored area. A custom dataset is trained to identify enemy soldiers so that the system can differentiate between normal movement and suspicious activity. Whenever an enemy soldier is detected, an alert is generated using a piezo buzzer with the help of an ESP32 WROOM module. Along with hardware alerts, a Python-based web dashboard is developed to allow remote monitoring. The dashboard provides live video streaming and detection information and includes basic features such as user login, registration, and forgot password option for secure access. By combining IoT hardware with deep learning-based detection, the proposed system reduces manual effort and improves response time. The system offers a simple, low-cost, and effective solution for improving border surveillance using modern technology.