← Back to Projects

Agriculture-Based Rover

AI & Computer Vision | Robotics | Raspberry Pi 4

X-Y Axis Scanning Machine Prototype

Overview

An AI-driven autonomous robot designed to revolutionize precision agriculture through targeted pesticide application. By integrating a hybrid CNN-Transformer architecture with BM3D image denoising, this system overcomes real-world field challenges like motion blur to achieve a remarkable 99.15% detection accuracy. Unlike traditional blanket spraying, this dual-view system analyzes infection severity in real-time to deliver precise micro-doses only where needed—massively reducing costs and promoting eco-friendly farming.

Key Features & Implementations

Intelligent Disease Analysis

Uses camera-based image analysis trained on a new plant disease dataset. The Hybrid CNN-Transformer architecture ensures robust, high-accuracy detection.

Dual-View Scanning System

Operates using a primary X-Y axis scanning machine for highly precise structural scanning, complemented by a secondary mobile Rover for dynamic field deployment.

BM3D Auto-Denoising

Incorporates a real-time auto-denoise function running efficiently on a Raspberry Pi 4 to clarify motion-blurred images during rover traversal.

Precision Spray & Future Weeding

Calculates the exact required pesticide dose for targeted micro-spraying. Future iterations propose a Laser Weed Remover for fully automated, chemical-free weeding.

Technologies Used

Python Raspberry Pi 4 C/C++ Hybrid CNN-Transformer Computer Vision BM3D Laser Systems (Proposed)