Google Earth Engine (GEE) cloud computing based crop classification using radar, optical images and Support Vector Machine Algorithm (SVM)

Successful crop classification process requires a reliable and efficient source of data and algorithms. Known classification algorithms or sources of data are not considered as definite solutions for improving the crop classification process due to the type of remote sensing data, and availability of field verification data. The existence of cloud computing technology services can help reduce the burden of retrieving, manipulating, processing, and validating big data. Google Earth Engine (GEE) is one of these technologies dedicated to spatial data processing. The remote sensing images, classification algorithm, and verification method are provided by the GEE cloud-computing platform. Assessing the potentialities of the recent new Sentinel satellite images in providing high spatial and temporal data is the main objective of this paper. A well-known algorithm Support Vector Machine Algorithm (SVM) is used to classify a series of Sentinel 1 (S1), Sentinel-2 A, and B (S2) images for different years. SVM algorithm is first used with Sentinel 2 data after tuning different needed parameters such as kernel's degree, gamma and cost. The SVM accuracy in the classification of Sentinel-2 images reached 93.6 %. When Sentinel-1 data (VH and VV bands) are combined with Sentinel-2 images, SVM accuracy increased to more than 96%.