This research aimed to develop a sensor system which is able to detect moving objects from a moving platform. Especially, now we are focusing to achieve it using a car as the moving platform in city area environments. There have been several research projects which achieved autonomous driving using multiple sensors and communications in a limited environment like highways. However, when we want to recognize much more complicated environment like city area, it requires much more ability to detect complicated environment including human who moves unexpectedly. Also, those things should be detected from a moving car which is also very difficult to detect its own motion because all sensors can not avoid certain amount of errors. In this research, we defined our target environment as city area which is rather clatter but man-made environment. Also, moving speed is considered slower than highway, and consequently target area of detection is limited as nearer field around the platform. Toward this problem, now we are focusing to develop methods using an omni-directional camera, vehicle odometry and simple range sensor like radar or sonar.
Platform: NavLab11Internal infomation: Images- Temp