Region-Level Motion- Based Background Modeling and Subtraction Using MRFs Shih-Shinh Huang Li-Chen...

Region-Level Motion-Based Region-Level Motion-Based BackgroundBackground

Modeling and Subtraction Modeling and Subtraction Using MRFsUsing MRFsShih-Shinh HuangShih-Shinh Huang

Li-Chen FuLi-Chen FuPei-Yung HsiaoPei-Yung Hsiao

2007 IEEE2007 IEEE

AbstractAbstract

This paper presents a new approach to aThis paper presents a new approach to automatic segmentation of foreground outomatic segmentation of foreground objects from an image sequence by integrbjects from an image sequence by integrating techniques of background subtracating techniques of background subtraction and motion-based foreground segmtion and motion-based foreground segmentation.entation.

OutlineOutline

INTRODUCTIONINTRODUCTION REGION-BASED MOTION REGION-BASED MOTION

SEGMENTATIONSEGMENTATION BACKGROUND MODELINGBACKGROUND MODELING MRFS-BASED CLASSIFICATIONMRFS-BASED CLASSIFICATION RESULTSRESULTS CONCLUSIONCONCLUSION

INTRODUCTIONINTRODUCTION

In many applications, success of detecting foreground regions from a static background scene is an important step before high-level processing.

In real-world situations, there exist several kinds of environment variations that will make the foreground detection more difficult.

Several kinds of environment variations

Illumination VariationGradual illumination variationSudden illumination variationShadow

Motion VariationGlobal motionLocal motion

System Overview

REGION-BASED MOTION REGION-BASED MOTION SEGMENTATIONSEGMENTATION

motion vector

Region Projection

Projecting regions in the previous frame to the current one, is to facilitate the segmentation.

Motion Marker Extraction

The output of this step is a set of motion-coherent regions, all pixels within a region comply with a motion model.

Boundary Determination

Merge uncertain pixels to one of the markers.

BACKGROUND MODELINGBACKGROUND MODELING

A brief description of Stauffer and Grimson’s work is first given and then we introduce the Bhattacharyya distance as the difference measure between the region from the region-based motion segmentation and the one represented by the background model.

Adaptive Gaussian Mixture Models

Bhattacharyya Distance

Shadow effectShadow effect

However, the region similarity defined in this way will lead to misclassification of the background region where direct light is blocked by the foreground object.

An example of shadow An example of shadow effecteffect

MRFS-BASED MRFS-BASED CLASSIFICATIONCLASSIFICATION

Incorporate the background model to classify every region in the segmentation map SM into either a foreground object or a background one by MRFs.

MRFs Framework

Region Classification

RESULTSRESULTS

CONCLUSIONCONCLUSION

Experimental results demonstrate that our proposed method can successfully extract the foreground objects even under situations with illumination variation, shadow, and local motion.

Our on-going research is to develop a tracking algorithm which can be used track the detected object.