1. Develpoment of Bagged Guava Quality Grading System Using Image Recognition and Generative Adversarial Networks(GANS)

Taiwan’s warm climate and abundant sunlight make it highly suitable for guava cultivation, making guava an important economic crop. However, current quality grading still relies on manual inspection, which is labor-intensive, inconsistent, and affected by bagging practices. To address this, we propose an automatic grading system using deep learning and generative adversarial networks. The framework collects images of bagged and bare guavas, applies YOLOv9 for fruit detection and background... C. Chang