PDR: A Performance Evaluation Method for Foreground-Background Segmentation Algorithms

TitlePDR: A Performance Evaluation Method for Foreground-Background Segmentation Algorithms
Publication TypeJournal Articles
Year of Publication2006
AuthorsKim K, Chalidabhongse TH, Harwood D, Davis LS
JournalEURASIP Journal on Applied Signal Processing
Date Published2006///
Abstract

We introduce a performance evaluation methodologycalled Perturbation Detection Rate (PDR) analysis for
measuring performance of foreground-background seg-
mentation. It has some advantages over the commonly
used Receiver Operation Characteristics (ROC) analy-
sis. Specifically, it does not require foreground targets
or knowledge of foreground distributions. It measures
the sensitivity of a background subtraction algorithm
in detecting possible low contrast targets against the
background as a function of contrast, also depending
on how well the model captures mixed (moving) back-
ground events. We compare four background subtrac-
tion algorithms using the methodology. The experimen-
tal results show how PDR is used to measure perfor-
mance with respect to detection sensitivity in interest-
ing low contrast regions.