14 · 2025 – 2026AIExperiments
Agricultural Computer Vision
Crop and Weed Classification for Smart Farming
Role & Context
Computer Vision Researcher
Student Research Project
Domain
Smart Agriculture & Robotics
Focus
Crop vs. Weed Classification
Target
Edge-Device Computer Vision
Application
Sustainable Farming
14 · Project OverviewStudent Research Project
An agricultural computer vision initiative exploring machine learning models to differentiate between cash crops and invasive weeds for targeted farming applications.
Project Context & Objectives
Over-spraying herbicides damages soil quality and incurs high costs. Computer vision models running on agricultural robots enable precision spraying directly onto weeds.
Key Features & Design Solutions
- Dataset curation of agricultural field imagery under diverse lighting conditions.
- Exploration of lightweight convolutional vision architectures suitable for edge devices.
- Evaluation of classification accuracy across crop varieties and weed species.
Key Deliverables
- Field Image Dataset & Classification Labels
- Edge-Compatible Model Benchmark Tests
- Precision Agriculture Exploration Documentation
Tools & Stack
Computer VisionPyTorchCrop/Weed ClassificationEdge AI