Deep learning (DL) is a powerful state-of-the-art technique for image processing including remote sensing (RS) images. This letter describes a multilevel DL architecture that targets land cover and crop type classification from multitemporal multisource satellite imagery. The pillars of the architecture are unsupervised neural network (NN) that is used for optical imagery segmentation and missing data restoration due to clouds and shadows, and an ensemble of supervised NNs. As basic supervised NN architecture, we use a traditional fully connected multilayer perceptron (MLP) and the most commonly used approach in RS community random forest, and compare them with convolutional NNs (CNNs). Experiments are carried out for the joint experiment of crop assessment and monitoring test site in Ukraine for classification of crops in a heterogeneous environment using nineteen multitemporal scenes acquired by Landsat-8 and Sentinel-1A RS satellites. The architecture with an ensemble of CNNs outperforms the one with MLPs allowing us to better discriminate certain summer crop types, in particular maize and soybeans, and yielding the target accuracies more than 85% for all major crops (wheat, maize, sunflower, soybeans, and sugar beet).

Many applied problems arising in agricultural monitoring and food security require reliable crop maps at national or global scale. Large scale crop mapping requires processing and management of large amount of heterogeneous satellite imagery acquired by various sensors that consequently leads to a “Big Data” problem. The main objective of this study is to explore efficiency of using the Google Earth Engine (GEE) platform when classifying multi-temporal satellite imagery with potential to apply the platform for a larger scale (e.g. country level) and multiple sensors (e.g. Landsat-8 and Sentinel-2). In particular, multiple state-of-the-art classifiers available in the GEE platform are compared to produce a high resolution (30 m) crop classification map for a large territory (~28,100 km2 and 1.0 M ha of cropland). Though this study does not involve large volumes of data, it does address efficiency of the GEE platform to effectively execute complex workflows of satellite data processing required with large scale applications such as crop mapping. The study discusses strengths and weaknesses of classifiers, assesses accuracies that can be achieved with different classifiers for the Ukrainian landscape, and compares them to the benchmark classifier using a neural network approach that was developed in our previous studies. The study is carried out for the Joint Experiment of Crop Assessment and Monitoring (JECAM) test site in Ukraine covering the Kyiv region (North of Ukraine) in 2013. We found that Google Earth Engine (GEE) provides very good performance in terms of enabling access to the remote sensing products through the cloud platform and providing pre-processing; however, in terms of classification accuracy, the neural network based approach outperformed support vector machine (SVM), decision tree and random forest classifiers available in GEE.

For many applied problems in agricultural monitoring and food security, it is important to provide reliable crop classification maps. Satellite imagery is extremely valuable source of data to provide crop maps in a timely way at moderate and high spatial resolution. Information on parcel boundaries that takes into account the spatial context may improve the quality of maps compared to pixel-based classification approaches. In general, parcels may contain several plots with different crops and such situations should be taken into account when using parcel boundaries. In this paper, we aim to compare pixel-based and parcel-based approaches to crop classification from multitemporal optical (Landsat-8) and synthetic-aperture radar (SAR) Sentinel-1 imagery. For this, we propose a parcel-based approach that involves a pixel-based classification map and specifically designed rules to account for several plots within parcel. The study is carried out for the Joint Experiment of Crop Assessment and Monitoring test site in Ukraine covering the Kyiv oblast (North of Ukraine) in 2013-2015, and the Odessa oblast (South of Ukraine) in 2014-2015. We found that pixel-based overall classification accuracy can be increased from 85.32% to 89.40% when using parcel boundaries. Among tested parcel-based approaches, the one that relied on pixel-based classification map and a procedure to select multiple plots within the parcel yielded the best performance.

Ukraine is one of the most developed agricultural countries in the world. For many applications, it is extremely important to provide reliable crop maps taking into account diversity of cropping systems used in Ukraine. The use of optical imagery only is limited due to cloud cover, and previous studies showed particular difficulties in discriminating summer crops in Ukraine such as maize, soybeans, sunflower, and sugar beet. This paper focuses on exploring feasibility and assessing efficiency of using multitemporal satellite synthetic-aperture radar (SAR) acquired in C-band and optical images for crop classification in Ukraine. Both optical (Landsat-8/OLI) and SAR (Radarsat-2) images are used to assess the impact of adding backscattering intensity from SAR images for classification purposes. SAR intensity information is very important due to availability of Sentinel-1 imagery over Ukraine starting March 2015. Different combinations of optical and SAR images, as well as SAR modes and polarizations, are assessed for better discrimination of crops. A committee of neural networks, in particular multilayer perceptrons (MLPs), is used to improve classification accuracy compared to several standard classifiers. It is found that using backscatter coefficients from SAR images alone provides the same performance for winter crops (wheat and rapeseed) as surface reflectance from optical images. Considering the summer crops, the major impact of adding backscatter intensity information from SAR images is in better separation of sunflower, soybeans, and maize.