engleski

Unity based Urban Environment Simulation for Autonomous Vehicle Stereo Vision Evaluation

Testing and evaluation of sensor processing algorithms for autonomous vehicles is challenging due to the problem of collecting reference data. To ensure safety and robustness, many man-hours need to be spent in collecting and preparing such data. One solution to alleviate this problem is to use computer simulations. Computer simulations can model a real system with all its static and dynamic characteristics. This approach provides efficiency and precision when collecting data and reduces testing time. The aim of this paper is development of a simulation environment based on Unity where it would be possible to test sensors and algorithms for autonomous vehicles and show deviations from reference data. The proposed simulation model contains typical city objects and participants: roads, sidewalks, buildings, pedestrians, traffic signs and vehicles. In this paper we simulate motion and sensors from a single vehicle equipped with a stereo camera setup. The program environment Unity is used for designing the simulation, and behavioral scripts are executed with C# programming language. To showcase the testing of applicable algorithms, OpenCV class for computing stereo correspondence, using the semi-global block matching algorithm, is used on simulated stereo images and we discuss future development of the simulation

Autonomous Vehicles ; Urban Scenario Simulation ; Stereo Vision

nije evidentirano