A Review on Land Cover Classification Techniques for Major Fruit Crops in India - Present Scenario and Future Aspects

Abstract

India is second largest fruit producer in the word. Mango (Mangifera indica L.) and banana (Musa sp.) are two major fruit crops grown in India. Fruit crops are very important for improving land productivity, economic condition of farmers by increasing income, generating employment and providing nutritional security. For better management of crops and bringing more area under fruit crops, the information on current status of crop production must be known. In brief literature review, it is observed that many researchers have used the satellite images for crop production by utilizing their surface reflectance or backscattering coefficient values. Classification of satellite images is the main source of information retrieval about the crop production and it also forms the basis and is an important step for crop cover classification, crop identification, acreage estimation, assessment of crop health and/or crop stress, change detection, yield prediction etc. There are several methods for satellite image classification such as ISODATA, K-Means, Maximum Likelihood, Minimum Distance, Artificial Neural Network (ANN), Support Vector Machine (SVM), and Decision Tree Classification (DTC), etc. Many researchers have used these methods for various purposes. The suitability/performance of these methods is also different and depends on the context of use, quality of imagery and ground truth data. In this paper, comparative suitability of unsupervised classifiers (ISODATA and K-Means), supervised classifiers (Parellepiped, Mahalanobis Distance, Maximum Likelihood and ANN) and Decision Tree Classification (DTC) for crop classification are reviewed.