Reference applications form a basis for designing and testing ADAS applications.
Simulation using realistic driving scenarios and sensor models is a crucial part of testing automated driving algorithms. Automated Driving Toolbox provides various options such as cuboid simulation environment, Unreal engine simulation environment, and integration with RoadRunner Scenario to test these algorithms. This application supports import and export of scenes and scenarios to ASAM OpenDRIVE and ASAM OpenSCENARIO® formats.
Ground Truth Labeling
Automate labeling of ground truth data and compare output from an algorithm under test with ground truth data. Using Ground Truth Labeler app, label multiple signals like videos, image sequences, and lidar signals representing the same scene.
Automate testing of ADAS algorithms and systems from requirements to implementation. Assess the functionality of automated driving applications (autonomous emergency braking, highway lane following, highway lane change) by defining scenarios and performing regression testing. Use scenario generation and variation tools to create scenarios from recorded sensor data and generate multiple variants from a seed scenario.
Planning and Control
Plan driving paths with vehicle costmaps and motion-planning algorithms. Use lateral and longitudinal controllers to follow a planned trajectory.
Detection and Tracking
Develop and test vision and lidar processing algorithms for automated driving. Perform multi-sensor fusion and multi-object tracking framework with Kalman filters.
Localization and Mapping
Use simultaneous localization and mapping (SLAM) algorithms to build maps surrounding the ego vehicle based on visual or lidar data. Access and visualize high-definition map data from the HERE HD Live Map service. Display vehicle and object locations on streaming map viewers.