Target Tracking using Radar and Image IoT Nodes

Abstract

The availability of multiple sensing and edge processing capabilities by sensing nodes, that may include radar, image, and acoustic, enables the collection of rich information to improve performance in target tracking. However, the tracking system, that includes a central data fusion center and IoT nodes, needs to extract information from heterogeneous data under constraints of IoT power, processing, and communication rates. This work presents a method for target tracking using image and radar data. The method is able to fuse heterogeneous data and direct agile edge processing of data on board of the sensing IoT nodes. Data fusion and agile edge processing improves tracking performance while reducing processing and communication rates. Simulation-based results, using synthetic radar data and real image data, demonstrate an improved tracking performance when using heterogeneous data versus using a single type of data.