• JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
 
  Bookmark and Share
 
 
Doctoral Thesis
DOI
https://doi.org/10.11606/T.45.2020.tde-19032020-094519
Document
Author
Full name
Joelma Cristina Costa e Silva
Institute/School/College
Knowledge Area
Date of Defense
Published
São Paulo, 2020
Supervisor
Committee
Silva, Flavio Soares Correa da (President)
Cybis, Helena Beatriz Bettella
Júnior, Orlando Fontes Lima
Leite, Daniel Furtado
Miranda, Paulo Andre Vechiatto de
Title in Portuguese
Análise baseada em contexto do movimento de pedestres em terminais de transporte
Keywords in Portuguese
Analise contextual
Aquisicao de dados de video
Comportamento de pedestres
Monitoramento de terminais de transporte
Abstract in Portuguese
A aquisicao de dados de deslocamentos de pessoas atraves de videos traz consigo uma imprecisao inerente que advem tanto da qualidade das imagens capturadas quanto da dinamica do movimento de pedestres no ambiente. A obtencao de informacoes precisas de rastreamento e contagem de pedestres a partir de videos ainda e um desafio. Este trabalho explora caracteristicas especificas do comportamento de pedestres no dominio dos terminais de transporte e apresenta uma ferramenta que agrega informacoes do contexto para aumentar a precisao dos dados obtidos de videos para a contagem de pedestres e determinacao dos fluxos em cenarios reais. A ferramenta proposta (1) rearranja os rotulos dos pedestres de acordo com o compor- tamento associado ao contexto considerado, (2) melhora a precisao na contagem dos pedestres a partir das rotas rastreadas, e (3) usa informacoes dinamicas de movimentacao para melhorar a identificacao do percurso completo realizado pelos pedestres individualmente. Com isso, a ferramenta proposta incrementa a acuracia das rotas rastreadas, agregando informacoes com base no comportamento caracteristico dos pedestres em terminais de transporte.
Title in English
Context-based understanding of pedestrian motion in transportation terminals
Keywords in English
Contextual analysis
Pedestrian behavior
Transport terminal monitoring
Video data-acquisition
Abstract in English
Acquisition of pedestrian data through video brings with it an inherent inaccuracy that comes from both the quality of the images captured and the dynamics of pedestrian movement in the environment. Getting accurate pedestrian tracking and counting information from videos is still a challenge. This paper explores specific pedestrian behavior characteristics in the transport terminals domain and presents a tool that aggregates context information, thus increasing the accuracy of data obtained from videos that count and determine pedestrian flows in real scenarios. The proposed tool (1) reorganizes pedestrian labels according to the context-associated behavior considered, (2) improves pedestrian counting accuracy from tracked routes, and (3) uses dynamic movement information to improve complete path identification taken by pedestrians individually. Consequently, the proposed tool increases the accuracy of the tracked routes by aggregating information based on the characteristic behavior of pedestrians in transport terminals.
 
WARNING - Viewing this document is conditioned on your acceptance of the following terms of use:
This document is only for private use for research and teaching activities. Reproduction for commercial use is forbidden. This rights cover the whole data about this document as well as its contents. Any uses or copies of this document in whole or in part must include the author's name.
teseFinalJoelma.pdf (8.06 Mbytes)
Publishing Date
2020-03-19
 
WARNING: Learn what derived works are clicking here.
All rights of the thesis/dissertation are from the authors
CeTI-SC/STI
Digital Library of Theses and Dissertations of USP. Copyright © 2001-2022. All rights reserved.