A5057033978- Remote Sensing in Agriculture
- Land Use and Ecosystem Services
- Remote Sensing and LiDAR Applications
- Smart Agriculture and AI
- Remote Sensing and Land Use
- Soil Geostatistics and Mapping
- Plant Water Relations and Carbon Dynamics
- Remote-Sensing Image Classification
- Advanced Image and Video Retrieval Techniques
- Forest ecology and management
- Geographic Information Systems Studies
- Species Distribution and Climate Change
- Spectroscopy and Chemometric Analyses
- Atmospheric and Environmental Gas Dynamics
- Tree-ring climate responses
- Advanced Neural Network Applications
- Multimodal Machine Learning Applications
- Urban Heat Island Mitigation
- Hydrological Forecasting Using AI
- Health, Environment, Cognitive Aging
- Human Pose and Action Recognition
- Domain Adaptation and Few-Shot Learning
- Greenhouse Technology and Climate Control
- Synthetic Aperture Radar (SAR) Applications and Techniques
- Video Analysis and Summarization
State of California
2018-2023
Descartes Labs (United States)
2019
University of California, Berkeley
2008-2016
For the two of most important agricultural commodities, soybean and corn, remote sensing plays a substantial role in delivering timely information on crop area for economic, environmental policy studies. Traditional long-term mapping corn is challenging as result high cost repeated training data collection, inconsistency image process interpretation, difficulty handling inter-annual variability weather progress. In this study, we developed an automated approach to map state Paraná, Brazil...
We report on a global cropland extent product at 30-m spatial resolution developed with two land cover maps (i.e. FROM-GLC, Finer Resolution Observation and Monitoring, Global Land Cover; FROM-GLC-agg) 250-m probability map. A common validation sample database was used to determine optimal thresholds of in different parts the world generate cropland/noncropland mask according classification accuracies for samples. decision tree then applied combine masks: one existing from literature other...
An accurate and timely crop-type map is essential in water planning California. So far, no effort has been made to effectively efficiently identify specific crop types on an annual basis this area. We have explored the potential of Moderate Resolution Imaging Spectroradiometer (MODIS) reflectance images annually major San Joaquin Valley, A phenology-based classification approach employed, which extracted phenological metrics from normalized difference vegetation index (NDVI) profiles...
Accurate predictions of wheat yields are essential to farmers'production plans and the international trade in wheat. However, only poor approximations productivity crops China can be obtained using traditional linear regression models based on vegetation indices observations yield. In this study, Sentinel-2 (multispectral data) ZY-1 02D (hyperspectral were used together with 15709 gridded yield data (with a resolution 5 m × m) predict winter These estimates four mainstream data-driven...
Mapping the corn dynamics at a large scale and multiple years is essential for global food security. Traditional mapping approaches by collecting training samples from field surveys are labor-intensive, challenging large-scale of over long term. This study developed an efficient approach to map in main districts United States (US) using adaptive strategies high-quality samples. First, this proposed automatic collect stable representative crop data layers (CDL) product. Then, improved...
Long-term time series of spatially explicit cropland maps are essential for global crop modelling and climate change studies. The spatial resolution temporal continuity have been improving several data sets released recently. Here, we calculated country-level areas from the annual land-cover (LC) produced by European Space Agency Climate Change Initiative (ESA-CCI) project Food Agricultural Organization United Nations statistical (FAOSTAT) 1992 to 2014. Because these two used different...
Abstract The goal of this study was to promptly map the extent corn and soybeans early in growing season. A classification experiment conducted for US Corn Belt neighboring states, which is most important production area world. To improve timeliness algorithm, training completely based on reference data images from other years, circumventing need finish collection current account interannual variability crop development cross-year scenario, several innovative strategies were used. random...
Bamboo forest is a unique landscape that mainly composed of herbal plants. It has stronger capability to increase terrestrial carbon sinks than woody forests in the same environment, thus playing special role absorbing atmospheric CO2. Accurate and timely bamboo maps are necessary better understand quantify their contribution hydrological cycles. Previous studies have reported phenology pattern forests, i.e., on- off-year cycle, can be detected with time-series high spatial resolution remote...
Abstract Modeling in computer vision has long been dominated by convolutional neural networks (CNNs). Recently, light of the excellent performance self-attention mechanism language field, transformers tailored for visual data have drawn significant attention and triumphed over CNNs various tasks. These heavily rely on large-scale pre-training to achieve competitive accuracy, which not only hinders freedom architectural design downstream tasks like object detection, but also causes learning...
It is not easy in remote sensing field to distinguish corn and soybean mapping for the similarity of mixed summer crops. To understand variations better generate maps more accurately, accurate these two crops required. However, classifying different with easy. Finding discriminating features use a classifier can lead higher-precision mapping. In this paper, we used feature selection random forests analyse most important phenological soybeans US Corn Belt (also called Extended Belt, ECB),...