Skip to content

Latest commit

 

History

History
10 lines (6 loc) · 975 Bytes

README.md

File metadata and controls

10 lines (6 loc) · 975 Bytes

Mushroom Vision 🍄

The computer vision for safe and informed knowledge about the gold of the mountains

This project was born from the combination of passion, technical skills and the need to have a tool that could help those who, like me, happened to be passionate about mushroom picking in the fall season.

The first project is Mushroom Vision 4 which classifies 4 species, 2 poisonous and 2 potentially edible respectively 2 amanites, muscaria and vaginata, and 2 boletus, edulis and erythropus with excellent results.

The next version will be able to work with 10 different species, Mushroom Vision 10.

An object detection model was built before Mushroom Vision 10 was implemented to overcome the problem of classifying photos that do not actually contain mushrooms into a particular species.