Research Projects
Explore research projects in digital agriculture, remote sensing, and precision farming conducted at UC Davis Digital Agriculture Laboratory.
📡 Sentinel-2 for Crop Yield Estimation
Open-Access Systematic Review of Field-Scale Yield Mapping from Satellite Time Series
Focus: Machine learning, crop-growth data assimilation, and Sentinel-1–Sentinel-2 fusion
Scope: Yield-centered synthesis of Sentinel-2 research since 2015 across crops and continents
Publication: Smart Agricultural Technology 14 (2026) 102405 · Open access CC BY
🛰️ AlphaEarth + Deep Learning for California Tomato Mapping
Field-Scale Processing Tomato Mapping with Geospatial Foundation-Model Embeddings
Technology: AlphaEarth 64-band embeddings, U-Net segmentation, AWS SageMaker, Monte Carlo dropout
Results: 99.04% F1 score and 98.11% IoU on 1,424 spatially independent test fields
Presentation: ASABE Annual International Meeting 2026 · Indianapolis, IN
🏆 Farm Robotics Challenge 2026
1st Rank, Excellence in Specialty Crops Award from Western Growers
Technology: Project BloomSense - UAV mapping fused with Amiga ground robot
Impact: High-resolution bloom density estimates for pollination and yield prediction
Achievement: Selected as winner among 96 global teams at USA Plug and Play Summit
🚁 Tomato Crop Monitoring with Drone LiDAR
Advanced Remote Sensing for Parasitic Weed Detection and Biomass Estimation
Technology: Matrice drone, LiDAR module, 3D point cloud analysis
Focus: Early detection of broomrape, biomass monitoring, stress indicators
Collaboration: Alireza Pourreza, Ali Moghimi, Mohsen Mesgaran
📷 Multispectral + Deep Learning for Broomrape Detection
Drone-Based Multispectral Imaging and LSTM for Timely Detection of Branched Broomrape in Tomato Farms
Technology: DJI Matrice 210, MicaSense Altum-PT, LSTM, SMOTE
Results: 88.37% accuracy, 95.37% recall; earliest detection at 897 GDD
Publication: SPIE 2024 (Vol. 13053) • Narimani, Pourreza, Moghimi, Mesgaran, et al.
🌿 Leaf Spectral Analysis for Broomrape Detection
Early Detection of Branched Broomrape in Tomato Using Leaf Reflectance and Ensemble Machine Learning
Technology: Portable spectroradiometer (350–2500 nm), RF, XGBoost, SVM, Naive Bayes
Results: 89% accuracy at 585 GDD; water absorption bands (1500, 2000 nm) key for early detection
Publication: IFAC-PapersOnLine 59(23), 2025 • Presented at AGRICONTROL 2025, Davis, CA
🛰️ Satellite Imagery and Time-Series for Broomrape Detection
Sentinel-2 and LSTM for Branched Broomrape Detection in Tomato Farms
Technology: Sentinel-2, 20 vegetation indices, 5 NN-derived traits, GDD alignment, LSTM
Results: 87% test accuracy; NDMI, CCC, FAPAR, CHL-RED-EDGE most influential
Publication: SPIE 2025 (Vol. 13475) • Orlando, FL • Narimani, Pourreza, Moghimi, et al.
🌱 Aeroponic Smart Greenhouse & Deep Learning
IoT-Controlled Aeroponic Greenhouse and AI-Based Plant Disease Detection
Technology: IoT (Arduino, Ubidots), centrifugal aeroponic irrigation, VGG-19, InceptionResNetV2, InceptionV3
Results: VGG-19 92% on industrial data, 86.34% on experimental greenhouse; healthy / drought / rust detection
Publication: ASABE 2021 Virtual Meeting · Paper 2101252 · Narimani, Hajiahmad, Moghimi, Alimardani, et al.
More Research Coming Soon
Additional research projects and publications in development