Home // MMEDIA 2012, The Fourth International Conferences on Advances in Multimedia // View article

A Q-Learning Approach to Decision Problems in Image Processing

Authors:
Alexandru Gherega
Monica Radulescu
Mihnea Udrea

Keywords: reinforcement learning; Q-Learning; computer cognition; adaptive and learning systems

Abstract:
Abstract—Decision making in a sequential and ncertain manner is still one of the major problems that all computer science research fields relate to, either by trying to find the optimal solution for it or needing such a solution to obtain optimal results for some other stated problem. This paper's objective is to address the possibility of using reinforcement learning for decision problems in data processing environments. A short review of current uses for reinforcement learning solutions into different fields is depicted as well as a particular case in image processing methods. A solution is proposed for a specific decision problem in the field of image processing. Our implementation shows the results of a reinforced parameterization method for edge detection using Q-Learning.

Pages: 60 to 66

Copyright: Copyright (c) IARIA, 2012

Publication date: April 29, 2012

Published in: conference

ISSN: 2308-4448

ISBN: 978-1-61208-195-3

Location: Chamonix, France

Dates: from April 29, 2012 to May 4, 2012