Analysis of Water Stress Using Sentinel-2 NDWI Images for Optimization of Progressive Harvest Dates in Blueberry (Vaccinium corymbosum) Crops: A Case Study of the Peruvian District of Nuevo Chao in the La Libertad Region †

Ríos C.; Machaca V.; Morales S.; Hernández S.; Angulo A. · Engineering Proceedings (2025) · DOI: 10.3390/2025083027

Resumen

This research proposes using the Sentinel-2 platform to cultivate blueberries in the district of Nuevo Chao, La Libertad, Peru, taking advantage of favorable climatic conditions such as soil moisture and its production. The results show a model performance of 93%, with a specific neural network configuration. The Grad-Cam tool enables the interpretation of field images, highlighting distinct features at different growth stages. The metrics analysis reveals a 93% classification accuracy in detecting the growth stages. The discussion emphasizes the importance of remote sensing for evaluating plantations, endorsing the effectiveness of Sentinel-2. This technology is essential for improving the management and efficiency regarding blueberry production in Nuevo Chao.

Instituciones: Universidad San Ignacio de Loyola