Using State-Space Database to Control Vehicles

Recent years, there are researches to operate vehicles by the voice instruction. It is important for these researches to recognize a peripheral environment. There are internal sensor and external sensor used to recognize a peripheral environment. Internal sensor is often used together with external sensor because the error with driving system is caused. Imaging sensor can especially measure the plane compared with other external sensors, the research combined with internal sensor is abundant. The delay is caused because it takes the processing time when imaging sensor is used. In this research, imaging sensor is installed on the vehicle, so the system is visual feedback control. The goal of this research is that using state-space database can control the vehicle. The vehicle can run on the straight corridor without the patterns like joint of the building. The vehicle cannot distinguish the patterns like joint of building on straight corridor and wall. Sobel operator that picks the edge causes this problem, because a similar feature appears to the objects. As result, the vehicle can be controlled regardless of the patterns like joint of the building on straight corridor. The state quantity of state-space database that solves this problem is found