Person re-identification using masked keypoints

Diego Reyes, John Atkinson

Producción científica: Capítulo del libro/informe/acta de congresoContribución a la conferenciarevisión exhaustiva

Resumen

In this work, a method for person re-identification from surveillance videos is proposed. In this approach, person detection is based on moving objects from sequences of images, and on incorporating a feature extraction technique that can distinguish distinct persons according to their physical appearance by using masked images that reduce noise from the background. Our approach uses keypoints to build an image’s descriptor so that the best discriminative keypoints can be identified between different persons. Experiments using our masked re-identification method show significant improvements in the recognition rate when masked frames are used to reduce noise of the second plane.

Idioma originalInglés
Título de la publicación alojadaRecent Trends and Future Technology in Applied Intelligence - 31st International Conference on Industrial Engineering and Other Applications of Applied Intelligent Systems, IEA/AIE 2018, Proceedings
EditoresOtmane Ait Mohamed, Malek Mouhoub, Samira Sadaoui, Moonis Ali
EditorialSpringer Verlag
Páginas45-56
Número de páginas12
ISBN (versión impresa)9783319920573
DOI
EstadoPublicada - 2018
Publicado de forma externa
Evento31st International Conference on Industrial, Engineering and Other Applications of Applied Intelligent Systems IEA/AIE 2018 - Montreal, Canadá
Duración: 25 jun. 201828 jun. 2018

Serie de la publicación

NombreLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volumen10868 LNAI
ISSN (versión impresa)0302-9743
ISSN (versión digital)1611-3349

Conferencia

Conferencia31st International Conference on Industrial, Engineering and Other Applications of Applied Intelligent Systems IEA/AIE 2018
País/TerritorioCanadá
CiudadMontreal
Período25/06/1828/06/18

Huella

Profundice en los temas de investigación de 'Person re-identification using masked keypoints'. En conjunto forman una huella única.

Citar esto