The proposed program aims to develop and validate nondestructive techniques for evaluating key quality attributes of guava fruit, including sweetness, acidity, and firmness, through the integration of image processing, near-infrared spectroscopy (NIRS), and machine learning (ML). By combining modern sensing technologies with data-driven modeling, this project seeks to establish rapid, reliable, and cost-effective approaches to assess guava quality without destructive sampling, thereby contributing to postharvest loss reduction and precision quality management in tropical fruit production systems.