Research Projects

There exists lack of research focused on utilizing DWR as locust swarm detection system. Even though DWR captures the signals from biological targets, identifying and demarcating them in the DWR images is highly laborious and requies skilled persons for interpretations. The increase in computation and the emerging methods in deep learning have shown the potential to address the time consuming human interpretation process. Applying the machine learning algorithms to detect locust swarms will reduce the time consuming extraction process. The present work proposes and advance machine learning algorithms for DWR quality control and for detecting the appearance of precipitation and biological targets in the Indian DWR products. For the proposed approach, DWR data from the Indian Meteorological Department is taken for selected case study dates to develop the framework of the proposed approach. The research hypothesis being tested is "Can DWR Quality control measures in conjunction with ML techniques be used to effectively predict the swarm of locusts?" The outcome of this project will provide a near real time system for detection of locusts using Doppler Weather Radar and Machine Learning approach, which will be important for agricultural management.