Vaibhav Kedia
B.Tech(Dual Degree) in Computer Science Engineering

Areas of Interest:Remote Sensing-Land Cover Classification, Pattern Recognition



Information of crop phenology is essential for evaluating crop productivity and crop management. Understanding the changes in the patterns of different vegetation species provide valuable information about the human activities on ecological environment and different cropping practice followed in an area. I work on creating efficient clustering algorithms for crop cover classification using time series of satellite imagery. Using the popular clustering techniques like K-Means to identify cluster centers and then clustering this cluster centers using DBSCAN, thus known as K-DBSAN, has provided better results than using them independently. Seasons play at important part in the cultivation of a crop. To incorporate the seasonal variations in identifying clusters K-DBSCAN is triggered in a hierarchical way.In the first step of the hierarchy, data is clustered using the Rabi season (”winter” season) EVI values. Each cluster obtained from the previous step is further clustered now using the Kharif season (”summer” season) EVI values. This hierarchy of clustering breaks down the patterns to seasonal level and clearly separates different cropping practices being followed.