souvik0306/Soil-Type-Classification

• Designed a classification model with 5 classes and 98% testing accuracy using a Convolutional Neural Network. • Applied Data Augmentation and reduced validation losses by 30% by applying MobileNetV2 Architecture.

28
/ 100
Experimental

This project helps classify soil types from images. You provide an image of soil, and it tells you if it's Black, Yellow, Cinder, Laterite, or Peat soil. This is useful for agriculturalists, geologists, or environmental scientists who need to quickly identify soil types for land management or research.

No commits in the last 6 months.

Use this if you need to identify one of five specific soil types from an image for educational purposes or as a starting point for more complex systems.

Not ideal if you need to classify a broader range of soil types or require highly accurate, production-ready results for critical applications, as the dataset is small.

agriculture geology soil-science land-management environmental-analysis
No License Stale 6m No Package No Dependents
Maintenance 0 / 25
Adoption 4 / 25
Maturity 8 / 25
Community 16 / 25

How are scores calculated?

Stars

7

Forks

6

Language

Jupyter Notebook

License

Last pushed

Oct 26, 2022

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/souvik0306/Soil-Type-Classification"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.