Application of fuzzy machine learning algorithm in agro-geography

Dual-sensor temporal satellite images help to fulfil required temporal dates to incorporate phenological/ sessional variation in the identification and mapping of specific vegetation/crop types. These types of information area were very much helpful in the agro-geography domain. This research focuses on the identification of different crops in a given area using temporal remote sensing data, using single as well as dual-sensor temporal Sentinel-1 and 2 satellite images based on a fuzzy modified possibilistic c-means classifier. The output assessment was conducted through Mean Membership Difference (MMD) method. MMD value was 0.007 within Mustard, between Mustard and Wheat was 0.077 and between Mustard and Grass was 0.101, which are the indicator of best separation within crops in a given area. This methodology is very much helpful to map specific crop spatially distributed in an area while using single or dual-sensor temporal data. Even within a crop-specific activity can also be mapped through this methodology and have great application in crop insurance.