Showing papers on "Skyline published in 1993"

Skyline For Video-based Virtual Rail For Vehicle Navigation

Abstract: A new concept of video-based virtual rail for vehicle navigation on pre- scheduled paths is proposed. The virtual rail uses skylines as landmarks for vehicle navigation in suburban residential areas. A video-based virtual rail is composed of skylines and the corresponding road locations. By matching the current skylines with the virtual rail, current vehicle location can be determined. Preliminary experimental results show that the longitudinal spatial resolution of one meter can be achieved.

Estimating ALV position in mountainous area

Abstract: A method of position estimation for an ALV (autonomous land vehicle) using features of the skyline is presented. The skyline from the vision system is assumed given and is compared with a computer map, called the CAD-MAP. The algorithm consists of identifying the peak points in the camera skyline, computing the ALV position for the identified peak points, and searching for the corresponding peak point in the CAD-MAP. Heuristics for computational efficiency and solution accuracy are also included in the algorithm. To test the validity and effectiveness of the algorithm, numerous simulations were performed and analyzed.

TAMPOPO: An Evolutionary Learning Machine Based on the Principle of Realtime Minimum Skyline Detection

Abstract: Summary A class of learning problems is formulated as a minimum skyline detection problem for an unknown, time-varying nonlinear function. As a solution to this problem, a constructive proof of the existence of a realtime minimum skyline detector called TAMPOPO is given, along with its formal specification in Concurrent Prolog. Its behavioral semantics, including its adaptability to both stable and unstable environments, are analyzed along the lines of fixed point theory. TAMPOPO is an open system with a memory of experience and a memory rewriting mechanism. It is assumed to have a prion knowledge-the Lipsitz constant-of a nonlinear function's topology. Each experience not resident on the minimum skyline is rewritten by a superior one nearer to the minimum-skyline. Residents on the minimum skyline constitute the fixed point of the rewriting mechanism. Moreover, the rewriting mechanism can catch up with fixed point changes by employing a form of primitive consciousness that detects and deletes specious experience through topological knowledge of the environment. These features make TAMPOPO more adaptable to time-varying unknown environments than traditional learning machines.