A Novel Approach for Freshness Detection in Vegetables Fruits Using Deep Learning and Principal Component Analysis
Natrayan L, N. Vijaya Anand, D. Srikar, S. Kaliappan, Ramya Maranan, Siva Kumar Pathuri
Abstract
Vegetable and Fruits freshness detection accuracy is crucial for maintaining quality control, cutting waste, and satisfying customer requests. By combining Principal Component Analysis (PCA) with the You Only Look Once (YOLO) deep learning framework, this study presents a unique method for freshness identification. Because of its real-time object identification capabilities, YOLO is used to extract features and locate objects precisely in high-resolution vegetable photos. In order to maximize computing efficiency while preserving crucial information, PCA is used to minimize the dimensionality of retrieved features. A large dataset of vegetables is used to test the suggested approach, which shows excellent accuracy and scalability in a variety of scenarios. According to experimental findings, YOLO and PCA combined offer a reliable, automated method for detecting freshness in real time. This strategy has a lot of promise to increase efficiency and quality control in retail settings and agricultural supply chains. And the proposed classifier is compared with base classifiers like Decision Tree and SVM in which the proposed classifier gave the best accuracy i.e., 96%.