The system is implemented using Raspberry Pi, camera module, Motion software, AWS cloud services, Mosquitto Broker, Postfix Mail Server, speaker module, and a mobile application developed using MIT App Inventor.
Working Principle
The system continuously monitors the agricultural field using a camera connected to Raspberry Pi. Motion software analyzes the live video feed and detects movement in the monitored area.
Experimental Setup of AI Agro Intrusion System
When motion is detected:
● The system captures images of the intrusion event.
● Raspberry Pi processes the captured data.
● Notifications are sent to the mobile application through Mosquitto Broker configured on AWS.
● Email alerts with captured images are sent using Postfix Mail Server.
● The speaker module generates warning sounds to scare away animals or intruders.
After completing the alert process, the system automatically returns to monitoring mode and continues surveillance.
Camera Monitoring Process
The camera module is connected to Raspberry Pi for continuous field surveillance.
Process:
1. Camera captures live video feed.
2. The video stream is continuously monitored by Motion software.
3. Monitoring continues in real-time during both day and night conditions.
Purpose:
● Continuous surveillance of agricultural fields
● Detection of unauthorized movement
● Image capturing during intrusion events
Motion Detection Process
Motion software is responsible for identifying movement in the monitored area.
Process:
1. Motion software compares video frames continuously.
2. If movement is detected beyond a defined threshold:
○ An intrusion event is triggered.
3. Captured images are stored temporarily for processing
Advantages:
● Automatic intrusion detection
● Reduced manual monitoring
● Real-time event processing
Notification and Alert Process
The system generates real-time alerts whenever intrusion is detected.
Mobile Notification Process:
1. Raspberry Pi sends intrusion data to Mosquitto Broker hosted on AWS.
2. MQTT protocol transfers notification messages.
3. Mobile application receives real-time alerts.
Email Alert Process:
1. The captured image is attached with an alert message.
2. Postfix Mail Server sends email to the farmer.
Benefits:
● Instant intrusion information
● Remote accessibility
● Fast response from farmers
Speaker Warning System
The speaker module is activated after intrusion detection.
Process:
1. Intrusion detected by Motion software.
2. Raspberry Pi sends a signal to the speaker module.
3. The speaker generates a warning sound.
Purpose:
● Scare away animals
● Alert nearby people
● Reduce crop damage
Mobile Application Implementation
The mobile application is developed using MIT App Inventor.
Features:
● Receives intrusion notifications
● Displays alert messages