Non-parametric background and shadow modeling for object detection

Tatsuya Tanaka, Atsushi Shimada, Daisaku Arita, Rin-Ichiro Taniguchi

Research output: Chapter in Book/Report/Conference proceedingConference contribution

20 Citations (Scopus)

Abstract

We propose a fast algorithm to estimate background models using Parzen density estimation in non-stationary scenes. Each pixel has a probability density which approximates pixel values observed in a video sequence. It is important to estimate a probability density function fast and accurately. In our approach, the probability density function is partially updated within the range of the window function based on the observed pixel value. The model adapts quickly to changes in the scene and foreground objects can be robustly detected. In addition, applying our approach to cast-shadow modeling, we can detect moving cast shadows. Several experiments show the effectiveness of our approach.

Original languageEnglish
Title of host publicationComputer Vision - ACCV 2007 - 8th Asian Conference on Computer Vision, Proceedings
Pages159-168
Number of pages10
EditionPART 1
Publication statusPublished - Dec 1 2007
Event8th Asian Conference on Computer Vision, ACCV 2007 - Tokyo, Japan
Duration: Nov 18 2007Nov 22 2007

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
NumberPART 1
Volume4843 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Other

Other8th Asian Conference on Computer Vision, ACCV 2007
CountryJapan
CityTokyo
Period11/18/0711/22/07

Fingerprint

Object Detection
Pixel
Pixels
Probability density function
Modeling
Density Estimation
Probability Density
Estimate
Fast Algorithm
Model
Range of data
Experiment
Background
Object detection
Experiments

All Science Journal Classification (ASJC) codes

  • Theoretical Computer Science
  • Computer Science(all)

Cite this

Tanaka, T., Shimada, A., Arita, D., & Taniguchi, R-I. (2007). Non-parametric background and shadow modeling for object detection. In Computer Vision - ACCV 2007 - 8th Asian Conference on Computer Vision, Proceedings (PART 1 ed., pp. 159-168). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 4843 LNCS, No. PART 1).

Non-parametric background and shadow modeling for object detection. / Tanaka, Tatsuya; Shimada, Atsushi; Arita, Daisaku; Taniguchi, Rin-Ichiro.

Computer Vision - ACCV 2007 - 8th Asian Conference on Computer Vision, Proceedings. PART 1. ed. 2007. p. 159-168 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 4843 LNCS, No. PART 1).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Tanaka, T, Shimada, A, Arita, D & Taniguchi, R-I 2007, Non-parametric background and shadow modeling for object detection. in Computer Vision - ACCV 2007 - 8th Asian Conference on Computer Vision, Proceedings. PART 1 edn, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), no. PART 1, vol. 4843 LNCS, pp. 159-168, 8th Asian Conference on Computer Vision, ACCV 2007, Tokyo, Japan, 11/18/07.
Tanaka T, Shimada A, Arita D, Taniguchi R-I. Non-parametric background and shadow modeling for object detection. In Computer Vision - ACCV 2007 - 8th Asian Conference on Computer Vision, Proceedings. PART 1 ed. 2007. p. 159-168. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); PART 1).
Tanaka, Tatsuya ; Shimada, Atsushi ; Arita, Daisaku ; Taniguchi, Rin-Ichiro. / Non-parametric background and shadow modeling for object detection. Computer Vision - ACCV 2007 - 8th Asian Conference on Computer Vision, Proceedings. PART 1. ed. 2007. pp. 159-168 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); PART 1).
@inproceedings{d1b9a414dd77444a969c5ecf7ecc2d1f,
title = "Non-parametric background and shadow modeling for object detection",
abstract = "We propose a fast algorithm to estimate background models using Parzen density estimation in non-stationary scenes. Each pixel has a probability density which approximates pixel values observed in a video sequence. It is important to estimate a probability density function fast and accurately. In our approach, the probability density function is partially updated within the range of the window function based on the observed pixel value. The model adapts quickly to changes in the scene and foreground objects can be robustly detected. In addition, applying our approach to cast-shadow modeling, we can detect moving cast shadows. Several experiments show the effectiveness of our approach.",
author = "Tatsuya Tanaka and Atsushi Shimada and Daisaku Arita and Rin-Ichiro Taniguchi",
year = "2007",
month = "12",
day = "1",
language = "English",
isbn = "9783540763857",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
number = "PART 1",
pages = "159--168",
booktitle = "Computer Vision - ACCV 2007 - 8th Asian Conference on Computer Vision, Proceedings",
edition = "PART 1",

}

TY - GEN

T1 - Non-parametric background and shadow modeling for object detection

AU - Tanaka, Tatsuya

AU - Shimada, Atsushi

AU - Arita, Daisaku

AU - Taniguchi, Rin-Ichiro

PY - 2007/12/1

Y1 - 2007/12/1

N2 - We propose a fast algorithm to estimate background models using Parzen density estimation in non-stationary scenes. Each pixel has a probability density which approximates pixel values observed in a video sequence. It is important to estimate a probability density function fast and accurately. In our approach, the probability density function is partially updated within the range of the window function based on the observed pixel value. The model adapts quickly to changes in the scene and foreground objects can be robustly detected. In addition, applying our approach to cast-shadow modeling, we can detect moving cast shadows. Several experiments show the effectiveness of our approach.

AB - We propose a fast algorithm to estimate background models using Parzen density estimation in non-stationary scenes. Each pixel has a probability density which approximates pixel values observed in a video sequence. It is important to estimate a probability density function fast and accurately. In our approach, the probability density function is partially updated within the range of the window function based on the observed pixel value. The model adapts quickly to changes in the scene and foreground objects can be robustly detected. In addition, applying our approach to cast-shadow modeling, we can detect moving cast shadows. Several experiments show the effectiveness of our approach.

UR - http://www.scopus.com/inward/record.url?scp=38149077423&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=38149077423&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:38149077423

SN - 9783540763857

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 159

EP - 168

BT - Computer Vision - ACCV 2007 - 8th Asian Conference on Computer Vision, Proceedings

ER -