Animal Detection and Alert System using Raspberry Pi and YOLO

Abstract

Human–animal conflict is a serious problem in rural and forest-adjacent areas, leading to crop damage and threats to human safety. This project presents a low-cost, edge-based intelligent system that detects the presence of animals within a predefined distance and identifies the type of animal using a deep learning model. The system uses a Raspberry Pi 4 and an ultrasonic sensor to detect proximity. Once an object is detected within 1 meter, a camera module is activated and a YOLO-based object detection model is used to classify the animal. Based on the detected animal (pig, cow, or elephant), a corresponding sound alert is played repeatedly to scare the animal away. The system continuously monitors the distance and stops operation when the animal leaves the area. Objectives

To detect the presence of animals within a specified distance using an ultrasonic sensor

To identify the type of animal using a deep learning-based object detection model (YOLO)

To generate audio alerts corresponding to the detected animal

To reduce unnecessary computation by activating the camera only when required

To develop a low-cost and energy-efficient edge AI solution

Hardware Requirements

Raspberry Pi 4 Model B (4GB recommended)

Ultrasonic Sensor (HC-SR04)

Raspberry Pi Camera Module

Speaker / Audio Output Device

Voltage Divider (for ECHO pin protection)

Power Supply (5V, 3A)

SD Card (32GB or higher)

Software Requirements

Raspberry Pi OS (64-bit)

Python 3

OpenCV

Ultralytics YOLOv5

Pygame

RPi.GPIO

PyTorch

System Methodology

The ultrasonic sensor continuously measures distance.

If an object is detected within 1 meter, the camera module is activated.

A single image frame is captured and passed to the YOLOv5 model.

The model detects and classifies the animal.

If the detected animal is pig, cow, or elephant, the corresponding sound is played 5 times.

The system rechecks the distance.

If the object is no longer within 1 meter, the system returns to idle state.

Block Diagram (Textual Description)

Ultrasonic Sensor → Raspberry Pi → Camera Module → YOLO Detection → Audio Alert System

Applications

Prevention of crop damage in agricultural fields

Early warning system in forest border areas

Wildlife monitoring systems

Smart farming solutions

Advantages

Low power consumption

Edge-based processing (no internet required)

Modular and scalable design

Real-time response

Limitations

Detection accuracy depends on lighting conditions

Limited to predefined animals

Performance constrained by Raspberry Pi hardware

Future Scope

Addition of more animal classes

Integration with GSM module for alerts

Use of thermal cameras for night detection

Cloud-based data logging and analytics

Hardware acceleration using Coral TPU

Conclusion

This project demonstrates the effective integration of embedded systems and artificial intelligence for real-world problem solving. By combining proximity sensing with deep learning-based object detection, the system minimizes power consumption while providing accurate and timely alerts. The project highlights the potential of edge AI in solving practical challenges in agriculture and wildlife management.

Developed By

Dhruv

Minor Project – Electronics and Communication Engineering (ECE) 6th semester