Large-Scale Crop Mapping From Multisource Remote Sensing Images in Google Earth Engine

Abstract

Large-scale crop mapping is vitally important to agriculrural monitoring and management. However, traditional methods cannot well meet the needs of large-scale applications. Therefore, this study proposed a method for large-scale crop mapping based on multisource remote sensing images. To be specific, 1) harmonic analysis was conducted on normalized difference vegetation index time-series derived from moderate resolution imaging spectroradiometer images and synthetic aperture radar backscattering coefficient time-series derived from Sentinel-1 data, respectively, extracting harmonic-derived phenological features and harmonic-derived backscattering features, and then combined with spectral features from Landsat-8 and Sentinel-2 images to construct the final multisource feature set for crop classification; 2) it employed prior constraints of crop dominance and cropland distribution to reduce misclassifications in large scale crop mapping; and 3) the whole process was conducted on the Google Earth Engine online platform, which can reduce the computational burdens caused by the spatiotemporal data. In the experimental study, we evaluated three crops, including wheat, rapeseed, and corn in Qinhai in 2018, based on the classification and regression tree classifier. The results show that the Jeffries-Matusita distances between crop samples are close to 2, and the overall accuracy is 84.25%. Furthermore, this study found that the distribution of the crops in Qinghai is associated with climate, topography, and cultivation habits.

Big Data and Machine Learning With Hyperspectral Information in Agriculture

“…An example of Big spectral data from satellite imagery is Sentinel-2. Sentinel-2 provides multispectral imaging (MSI) functionalities with spatial, spectral and temporal resolutions, and also has two spectral bands in the red-edge region for distinguishing the different agricultural crops [14]. Table I shows a summary of satellites and its hyperspectral/multispectral data capabilities from different countries in the world.…”

“…The authors in [14] proposed a large-scale crop mapping from multisource remote sensing images in Google Earth Engine. There are three stages in their approach: (1) Harmonic analysis on NDVI data combined with spectral features obtained from satellites (Landsat-8 and Sentinel-2); (2) Utilizing prior constraints of crop distribution and dominance; and (3) Information processing with Google Earth Engine.…”

A Gaussian Kernel-Based Spatiotemporal Fusion Model for Agricultural Remote Sensing Monitoring

“…Then the Savitzky-Golay (S-G) filter was performed on it. To improve the crop mapping accuracy, fused NDVI was firstly masked by the existing cropland product [37]. Because most corn and soybeans were cultivated at the mountain's foot, the planting structure is scattered.…”

Distributed Fusion of Heterogeneous Remote Sensing and Social Media Data: A Review and New Developments

“…There has also been a significant amount of research based on multisource remote sensing data for other applications. In [38], phenological features and backscattering features generated from Sentinel-1 images were combined with spectral features extracted from Sentinel-2 and Landsat 8 OLI images to construct multisource feature sets for crop mapping. The work presented in [39] described that freely available multisource data (e.g., Landsat TM, ETM+ imagery, JERS-1 active radar L-band imagery, and elevation data) contributed to a tree bagging classification procedure for wetland mapping.…”