Smart Disease Predictor for plants.

Making disease prediction easier and prevent plants from further rotting.

Comment

Inspiration

Imagine you found something fishy with your plant leave and you don't want to call for help and find the problem by yourself.

What it does

This project can detect plant diseases (here I have only made for 1 plant species i.e Tomato plant) and tell you the name of disease.

How we built it

Architecture used

  1. VGG19
  2. InceptionV3
  3. No. of classes: 10
  4. Classes :- ['Tomato_Late_blight' , 'Tomatohealthy' , 'TomatoEarly_blight' , 'TomatoSeptoria_leaf_spot' , 'TomatoTomato_Yellow_Leaf_Curl_Virus' , 'TomatoBacterial_spot' , 'TomatoTarget_Spot' , 'TomatoTomato_mosaic_virus' , 'TomatoLeaf_Mold' , 'Tomato_Spider_mites Two-spotted_spider_mite']
  5. Metrics for 10 Epochs (for VGG19) loss: 0.0531 accuracy: 0.9804 val_loss: 0.2975 val_accuracy: 0.9210
  6. Metrics for 10 Epochs (for InceptionV3) loss: 0.3708 accuracy: 0.9639 val_loss: 4.1287 val_accuracy: 0.8198

Challenges we ran into

Making the prediction accurate was pretty difficult and we had to try out multiple DEEP LEARNING algorithms to find better accuracy.

Accomplishments that we're proud of

Successfully able to predict the disease and help to take immediate actions.

What we learned

We can solve all the disease issues as early as possible and prevent further loss of crop/plant production.

What's next for Smart Disease Predictor for plants.

We can add more dataset and we will be able to predict all the plant related diseases in near future.

Built With

Share this project:

Updates