Note: This content is accessible to all versions of every browser. However, this browser does not seem to support current Web standards, preventing the display of our site's design details.


An ADP formulation with radial basis functions for camera surveillance


A. Martinez Gomez

Semester Thesis, HS13 (10310)

An autonomous surveillance system, consisting of evaders (in our case robots), and pursuers (in our case cameras), can be analyzed for various tasks. The applications of such a system, if solved efficiently, are commercially very appealing. Such an autonomous surveillance system could for example track athletes in a TV show or sport event, or it might track cars in a parking lot. A stochastic reachability formulation for this kind of autonomous surveillance problems was recently developed at IfA. Solutions via Dynamic Programming to this problems have been already implemented at the institute, but they are computationally very demanding. In our attempt to push scalability of the solution of the proposed formulation, an Approximate Dynamic Programming (ADP) technique using Linear Programming with a class of radial basis functions was adapted to the existing surveillance framework. On the project we investigated this later approach from both a theoretical standpoint and a computational one, especially by comparing different methods to calculate the probability of whether a specific camera state captures the evader. The goal of this project was to formulate some autonomous tasks in the ADP technique for the reachability framework, and design a simulation platform to investigate the applicability of the method.


Type of Publication:

(13)Semester/Bachelor Thesis

N. Kariotoglou

File Download:

Request a copy of this publication.
(Uses JavaScript)
% Autogenerated BibTeX entry
@PhdThesis { Xxx:2014:IFA_4756
Permanent link