Convolutional Neural Networks for automated recognition of plant leaves

About the project

Project for the module Algoritms for Massive Data of the course Algoritms for Massive Data, Cloud and Distributed Computing. The task is to implement a deep-learning-based system classifying pictures of leaves according to the twelve types of plants using the «Plant leaves» dataset. The «Plant leaves» dataset is published on Kaggle and released under the Community Data License Agreement – Sharing, Version 1.0. Here the data: https://www.kaggle.com/datasets/csafrit2/plant-leaves-for-image-classification.

Implemented CNNs models

• Binary classification (healthy-diseased)
• Multiclass classification (12 species)
• Multiclass classification (22 complete names)

Built With

• Python 3.10.11
• Tensorflow version: 2.9.2
• OpenCV

Composition of the repository

• Project_Plant_Leaves_Magnani.jpynb: this file containins libraries, data download, data preprocessing, CNNs architecture, hyperparameter tuning, regularization.
• Report_Leaves_Magnani.pdf: LateX report containing all the theoretical details about the implemented CNNs and the comparable results.