Micro Doppler Radar Data from Road Users

Micro-dopper radar data from road users to train neural networks. A data file contains time-synchronous radar and video data. A dataset includes radar data (distance, direction and speed) of road users recorded in a cross-section area in different traffic situations. In addition, the micro-doppler data and the video image of the traffic scene are stored for each moving radar target together with a time stamp. The micro-doppler data can be used to train neural networks after the user has assigned this data a label from certain classes. This can be done manually or automatically with the help of image recognition tools. Useful classes can be for this data:Person walking, person running, cyclist, vehicle, etc. The labeled data can then be used to train or verify neural networks. The aim of this classification is to identify traffic situations at an early stage, which may lead to a risk, in particular for vulnerable road users, in order to be able to transmit information or warnings to road users in a timely manner or to influence traffic management. In the future, this will also support autonomous driving in urban intersections.