A Comprehensive Analysis of Classic Machine Learning and Deep Learning Modeling for Breed Recognition of Bananas

This project presents an in-depth analysis of machine learning and deep learning models for the breed recognition of bananas. The dataset used consists of images of various banana varieties from Bangladesh, categorized into four different classes: Sabri Kola, Sagor Kola, Bangla Kola, and Champa Kola. The main objective is to apply both classic machine learning and advanced deep learning algorithms for breed classification, comparing their effectiveness and performance. The deep learning models utilized include EfficientNet, VGG11, DenseNet121, and VGG19, while the machine learning algorithms involved are LightGBM, CatBoost, and AdaBoost. The classification system is implemented in a user-friendly web application, allowing users to upload images, register, and classify the banana varieties efficiently. The system aims to assess the performance of these models, focusing on accuracy, computation time, and scalability. The goal of the project is to provide a reliable and effective solution for banana breed recognition, benefiting the agricultural sector and improving the processes involved in banana cultivation, sorting, and distribution.

Keywords: Banana classification, Machine learning, Deep learning, EfficientNet, VGG11, DenseNet121, VGG19, LightGBM, CatBoost, AdaBoost.