Validation of Object Identification and Tracking in Radar Imagery

Daniel T Dietz, Purdue University


Analysis of storm cells and tracking thereof in radar imagery requires substantial effort if done manually. An automated technique for performing cell identification and tracking can greatly ease the burden of performing the work by hand. For analyzing post-event cases automated techniques are a time-saver for the investigators and when analyzing real-time radar data or NWP output, applying automated techniques is necessary because the products must be made available in real-time to be useful. Evaluating the performance of such techniques is also a time consuming task because the comparative truth must be found by hand. A method of simulating storm cells so that the exact truth is known is explored as an alternative to evaluating identification and tracking algorithms against hand-identified cases.




Baldwin, Purdue University.

Subject Area


Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server