A RAM based neural network approach to people counting

Andrew J Schofield, T J Stonham, Pratap A Mehta

Research output: Chapter in Book/Published conference outputConference publication

Abstract

The designers of large buildings such as office blocks, shopping centres, and railway stations need to be able to predict building usage patterns. To do this they may conduct manual surveys in existing buildings. For example. surveys of lift usage in one building can be used to design the lift system for other buildings of a similar nature. Unfortunately, building usage surveys are labour intensive and can only be conducted over limited time periods. An automated method for people counting could reduce the cost and increase the utility of such surveys. There is an additional and growing requirement among the operators of large buildings for systems which can continuously monitor the number of people in an area. This information might then be used to limit overcrowding, or to determine the allocation of building services. This study concentrates on the use of people counting for a lift control application.
Original languageEnglish
Title of host publicationFifth International Conference on Image Processing and its Applications, 1995.
Place of PublicationLondon
PublisherIEEE
Pages652-656
ISBN (Print)0-85296-642-3
DOIs
Publication statusPublished - 1995

Fingerprint

Dive into the research topics of 'A RAM based neural network approach to people counting'. Together they form a unique fingerprint.

Cite this