Unsupervised dense crowd detection by multiscale texture analysis

Antoine Fagette; Nicolas Courty; Daniel Racoceanu; Jean Yves Dufour

doi:10.1016/j.patrec.2013.09.020

Unsupervised dense crowd detection by multiscale texture analysis

Antoine Fagette, Nicolas Courty, Daniel Racoceanu, Jean Yves Dufour

Research output: Contribution to journal › Article › peer-review

21 Scopus citations

Abstract

This study introduces a totally unsupervised method for the detection and location of dense crowds in images without context-awareness. With the perspective of setting up fully autonomous video-surveillance systems, automatic detection and location of crowds is a crucial step that is going to point which areas of the image have to be analyzed. After retrieving multiscale texture-related feature vectors from the image, a binary classification is conducted to determine which parts of the image belong to the crowd and which to the background. The algorithm presented can be operated on images without any prior knowledge of any kind and is totally unsupervised.

Original language	English
Pages (from-to)	126-133
Number of pages	8
Journal	Pattern Recognition Letters
Volume	44
DOIs	https://doi.org/10.1016/j.patrec.2013.09.020
State	Published - 15 Jul 2014
Externally published	Yes

Keywords

Dense crowd
Diffusion maps
Feature extraction
Multiscale
Quadtree
Segmentation

Access to Document

10.1016/j.patrec.2013.09.020

Cite this

@article{e21e54a4c65042b0877395f835ec2dd7,

title = "Unsupervised dense crowd detection by multiscale texture analysis",

abstract = "This study introduces a totally unsupervised method for the detection and location of dense crowds in images without context-awareness. With the perspective of setting up fully autonomous video-surveillance systems, automatic detection and location of crowds is a crucial step that is going to point which areas of the image have to be analyzed. After retrieving multiscale texture-related feature vectors from the image, a binary classification is conducted to determine which parts of the image belong to the crowd and which to the background. The algorithm presented can be operated on images without any prior knowledge of any kind and is totally unsupervised.",

keywords = "Dense crowd, Diffusion maps, Feature extraction, Multiscale, Quadtree, Segmentation",

author = "Antoine Fagette and Nicolas Courty and Daniel Racoceanu and Dufour, {Jean Yves}",

year = "2014",

month = jul,

day = "15",

doi = "10.1016/j.patrec.2013.09.020",

language = "English",

volume = "44",

pages = "126--133",

journal = "Pattern Recognition Letters",

issn = "0167-8655",

}

TY - JOUR

T1 - Unsupervised dense crowd detection by multiscale texture analysis

AU - Fagette, Antoine

AU - Courty, Nicolas

AU - Racoceanu, Daniel

AU - Dufour, Jean Yves

PY - 2014/7/15

Y1 - 2014/7/15

N2 - This study introduces a totally unsupervised method for the detection and location of dense crowds in images without context-awareness. With the perspective of setting up fully autonomous video-surveillance systems, automatic detection and location of crowds is a crucial step that is going to point which areas of the image have to be analyzed. After retrieving multiscale texture-related feature vectors from the image, a binary classification is conducted to determine which parts of the image belong to the crowd and which to the background. The algorithm presented can be operated on images without any prior knowledge of any kind and is totally unsupervised.

AB - This study introduces a totally unsupervised method for the detection and location of dense crowds in images without context-awareness. With the perspective of setting up fully autonomous video-surveillance systems, automatic detection and location of crowds is a crucial step that is going to point which areas of the image have to be analyzed. After retrieving multiscale texture-related feature vectors from the image, a binary classification is conducted to determine which parts of the image belong to the crowd and which to the background. The algorithm presented can be operated on images without any prior knowledge of any kind and is totally unsupervised.

KW - Dense crowd

KW - Diffusion maps

KW - Feature extraction

KW - Multiscale

KW - Quadtree

KW - Segmentation

UR - http://www.scopus.com/inward/record.url?scp=84901192326&partnerID=8YFLogxK

U2 - 10.1016/j.patrec.2013.09.020

DO - 10.1016/j.patrec.2013.09.020

M3 - Article

AN - SCOPUS:84901192326

SN - 0167-8655

VL - 44

SP - 126

EP - 133

JO - Pattern Recognition Letters

JF - Pattern Recognition Letters

ER -

Unsupervised dense crowd detection by multiscale texture analysis

Abstract

Keywords

Access to Document

Other files and links

Fingerprint

Cite this