Home // ICONS 2012, The Seventh International Conference on Systems // View article

Intelligent Processing of Video Streams for Visual Customer Behavior Analysis

Authors:
Johannes Kröckel
Freimut Bodendorf

Keywords: Customer tracking; video analysis; behavior analysis; retail; point of sale.

Abstract:
In today’s society purchasing goods through web shops has become habitual. Some years ago only a few products like books, computer games and music CDs were intensely sold by online retailers. Today’s internet shops are offering almost every imaginable product and service. This also leads to an increasing competition for traditional retailers offering products in stationary retail stores. Losing more and more customers stationary retailers need to think of new approaches for customer retention. Since customer retention is based on knowledge about the customers and their behavior store managers have to come up with new concepts for gaining and using customer knowledge to compete with bargain prices and 24/7 availability. In order to gain this knowledge without using vague customer surveys or short-time observations an automated solution is desirable. In this paper an approach is introduced, which allows to track and analyze customer movements through the store. Person tracking is accomplished by using aerial mounted cameras and a set of computer vision algorithms. Based on the captured movement data customer behavior analysis is performed by applying the dbscan algorithm and Markov models. The approach is illustrated by a test environment showing considerable differences in customer behavior for two settings.

Pages: 163 to 168

Copyright: Copyright (c) IARIA, 2012

Publication date: February 29, 2012

Published in: conference

ISSN: 2308-4243

ISBN: 978-1-61208-184-7

Location: Saint Gilles, Reunion

Dates: from February 29, 2012 to March 5, 2012