Discrimination of cash and grain crops using SVM classifier-an attempt on sentinel 1

The study aims to discriminate crops in to healthy and stressed categories of cash crops and grain crops by the support of machine language technique called support vector machine (SVM) with soil moisture and plant water content (PWC) as inputs. The soil moisture was arrived from scatter plot model [SP model] and PWC was retrieved from inverting the existing water cloud model (WCM). The crops were classified under the categories of Grain crops stressed (0 - 0.15), Grain crops healthy (0.16 - 3), Cash crops stressed (0.31 - 0.45) and Cash crops healthy (0.46 - 0.60) using SVM. Further, the results from SVM were validated using accuracy assessment method using the field observations. The outcomes demonstrate a high degree of agreement with the field circumstances (OA=85%, Kappa coefficient =0.83).