Abstract:
|
Driver Assistance Systems are in-vehicle systems that improve security in the automobile and increase comfort while driving. They have been acquiring great importance in recent years and will be very relevant in the near future. Continental develops this kind of product for different OEMs (Original Equipment Manufacturer) from the automobile sector. A series of sensors need to be mounted around the vehicle to capture information about the environment, which is stored in order to be processed and used. The purpose of this study is to investigate which is the best method for storage of occupancy grid maps captured by sensor systems mounted on vehicles. Two methods useful for compressing occupancy grid maps were implemented and evaluated in Matlab: the Wavelet Transform and the Quadtree algorithm. Then, these methods were tested to measure their rate of memory saving. At this stage the results revealed that the Quadtree algorithm works best. In a second stage, the Quadtree algorithm was implemented in ANSI-C so as to be embedded into the current code which simulates the controller environment. With the results obtained in this thesis, the memory release, thanks to the Quadtree algorithm, may be used to store more information about the environment, so that the current functions can be enhanced and the security when driving improved. |