/Bee-Hive-Health-Analyzer

Ai Powered Web App to determine bee hive health by analyzing a bee image - Tensorflow.js

Primary LanguageJupyter Notebook

Bee-Hive-Health-Analyzer

Live Web App: http://bee.test.woza.work/



This is a prototype for a freely available online tool that can predict the health of a bee hive by looking at an image of a bee. This could help bee keepers and bee researchers evaluate the health of a bee hive without having to disturb the hive itself.

The app outputs 6 possible bee hive conditions:

0: healthy
1: few varrao, hive beetles
2: varroa, small hive beetles
3: ant problems
4: hive being robbed
5: missing queen

Model F1 score: 0.94

The process used to build and train the model is described in this Kaggle kernel:
https://www.kaggle.com/vbookshelf/bee-hive-health-analyzer-web-app

The dataset used for the training can be found here:
https://www.kaggle.com/jenny18/honey-bee-annotated-images

All javascript, html and css files used to create the web app are available in this repo.