vsrekha213/leafdoctor

LeafDoctor - Detect Diseases in Plant Leaf

The objective of this project is to identify the type of diseases in the plants, so that it can be treated as soon as they appear.

Data Set can be found here : https://www.kaggle.com/emmarex/plantdisease

Data Set Description

The data set consists of 14 different diseases of the plants like pepper bell bacterial, pepper bell healthy, toamato late blight, early blight etc...

Model Training

Model architecture used is as follows :

1. Feature Extraction

1. Convolution(32 -> 3 X 3)
2. Normalization
3. Pooling(Max)
4. Dropout
5. Convolution(32 -> 3 X 3)
6. Normalization
7. Pooling(Max)
8. Dropout

2. Classification

1. Dense Layer (1024)
2. Activation used here is # Softmax

How to Use this Source Code

1. clone this repo
2. First install required tf packages and before that ensure to create separate conda environment and make sure tf version is 2.2
3. Run app.py file
4. To do prediction download 'model.h5' {you can directly uses this file for prediction}