2024 - Vegetable Scanner

In this project, I built a fruit and vegetable image classifier using machine learning techniques. I extracted visual features from each image using HOG, HSV histograms, LBP, and GLCM methods, and trained a Random Forest classifier to distinguish between 16 different classes. The model achieved around 64% accuracy and was evaluated using metrics like precision, recall, and F1-score.

To make the system interactive, I integrated the model into a PHP-based web interface. Users can upload an image and receive instant predictions about which fruit or vegetable it represents. This project demonstrates the application of computer vision in real-world classification tasks, with a focus on accessibility through a web-based interface.

User gets an instant prediction of the corresponding fruit or vegetable. At the end of each prediction, the system displays a full confusion matrix alongside the class labels, giving users detailed insight into model performance across all 16 categories.