AI Education
Computer Vision for Agriculture Demo
A GDIW 2025 workshop notebook connecting computer-vision fundamentals with face recognition and crop-disease use cases.
View on GitHub
Computer Vision & Facilitation
Project overview
Created to facilitate a practical workshop, this notebook introduces AI, machine learning, deep learning, convolution, and image representation before moving into hands-on CNN work and an agriculture-focused tomato-leaf demonstration.
The problem
Learners often encounter computer vision as abstract theory without a clear bridge from image pixels and convolution to real applications in agriculture and other sectors.
Approach
- 1
Teach the relationship between AI, machine learning, deep learning, and computer vision.
- 2
Demonstrate how images are represented numerically and processed through convolution.
- 3
Train and evaluate a CNN using the Olivetti Faces dataset.
- 4
Use MobileNetV2 to demonstrate feature extraction from uploaded tomato-leaf images.
- 5
Discuss how a crop-specific dataset such as PlantVillage would be needed to train a production disease classifier.
Outcomes and insights
Delivered a reusable Google Colab learning resource for the GDIW 2025 workshop.
Connected model generalization and confidence to practical prediction behavior.
Presented agriculture and healthcare as applied computer-vision pathways.