1 The MPEG-7 Visual Standard for Content Description-An Overview Thomas Sikora, Senior Member, IEEE A presentation by Modupe Omueti For CMPT 820:Multimedia.

11

The MPEG-7 Visual Standard for The MPEG-7 Visual Standard for Content Description-An OverviewContent Description-An Overview

Thomas Sikora, Thomas Sikora, Senior Member, IEEESenior Member, IEEE

A presentation by Modupe OmuetiA presentation by Modupe Omueti

ForFor

CMPT 820:Multimedia Systems CMPT 820:Multimedia Systems Spring 2005Spring 2005

22

ContentsContents

IntroductionIntroductionScopeScopeMethodologyMethodologyVisual DescriptorsVisual DescriptorsConclusionConclusion

33

IntroductionIntroduction

Moving Pictures Expert GroupMoving Pictures Expert GroupMPEG-1 for interactive video (1992)MPEG-1 for interactive video (1992)MPEG-2 for digital television (1994)MPEG-2 for digital television (1994)MPEG-4 for multimedia with emphasis on MPEG-4 for multimedia with emphasis on

visual objects (1998 v1, 1999 v2)visual objects (1998 v1, 1999 v2)MPEG-7 for multimedia content description MPEG-7 for multimedia content description

(2001)(2001)

44

TrendsTrends

Initially few sources of audio, image and Initially few sources of audio, image and videovideo

Increase in volume of digitized audio, Increase in volume of digitized audio, images and videoimages and video

Still images digital videoStill images digital video

55

MPEG-7MPEG-7

Formally named Multimedia Content Description Formally named Multimedia Content Description InterfaceInterface

Supports some degree of interpretation of the Supports some degree of interpretation of the information’s meaninginformation’s meaning

Interpretation can be passed on to or accessed Interpretation can be passed on to or accessed by a device or computer codeby a device or computer code

Not aimed at one application in particularNot aimed at one application in particular

66

ScopeScope GoalsGoals

Standardized descriptionsStandardized descriptions Meaningful descriptionsMeaningful descriptions

ElementsElements Description tools: visual decriptors and description Description tools: visual decriptors and description

schemesschemes Description Definition LanguageDescription Definition Language System toolsSystem tools

Figure 1: Scope of MPEG-7 Figure 1: Scope of MPEG-7 Figure 2: MPEG-7 main elementsFigure 2: MPEG-7 main elements

77

Figure 1: Scope of MPEG-7Figure 1: Scope of MPEG-7

Normative part of MPEG-7 standard

88

Figure 2: MPEG-7 main elementsFigure 2: MPEG-7 main elements

99

ApplicationsApplicationsDigital libraries (image catalogue, film)Digital libraries (image catalogue, film)Broadcast media selection (TV Broadcast media selection (TV

channels)channels) Investigation services (human Investigation services (human

characteristics recognition, forensics)characteristics recognition, forensics)Multimedia editing (personalized Multimedia editing (personalized

electronic news service)electronic news service)Figure 3: Abstract RepresentationFigure 3: Abstract Representation

1010

Figure 3: Abstract representation of Figure 3: Abstract representation of possible applications using MPEG-7possible applications using MPEG-7

1111

MethodologyMethodology

Standard Development

Specification for Technology Requirements

Technology Request

Proposal Evaluation

Experimentation Model Definition

Core Experiments

1212

Visual DescriptorsVisual Descriptors

General visual descriptors General visual descriptors Color, texture, shape, and motion featuresColor, texture, shape, and motion features

Domain specificDomain specific Identification of human faces and face Identification of human faces and face

recognitionrecognition

1313

Visual Color Visual Color DescriptorsDescriptors Color Spaces (HSV, HMMD)Color Spaces (HSV, HMMD)

Supports above for normative purposes Supports above for normative purposes Also supports RGB, YCbCr color spacesAlso supports RGB, YCbCr color spaces

Scalable color descriptorScalable color descriptor Figure 4Figure 4 Global color Distribution of Images in color Global color Distribution of Images in color

histograms histograms HSV space, uniformly quantized into 255 bins HSV space, uniformly quantized into 255 bins Haar Transform used to encode histogram Haar Transform used to encode histogram Histogram bin non-uniformly quantized Histogram bin non-uniformly quantized color coefficients or histogram bin values for color coefficients or histogram bin values for

matchingmatching

1414

Visual Color DescriptorsVisual Color Descriptors Dominant color descriptorDominant color descriptor

Global + local spatial color distributionGlobal + local spatial color distribution Colors clustered into a small no of representative colors Colors clustered into a small no of representative colors

representative color, %age, spatial coherency, variance representative color, %age, spatial coherency, variance Color layout descriptorColor layout descriptor

Spatial distribution of color in an arbitrarily shaped regionSpatial distribution of color in an arbitrarily shaped region Color structure descriptorColor structure descriptor

HMMD, local color feature, sliding windowHMMD, local color feature, sliding window Histogram on color appearance countHistogram on color appearance count

Group of Frames/Group of PicturesGroup of Frames/Group of Pictures SCD for a collection of similar images (frames) or video SCD for a collection of similar images (frames) or video

framesframes Average, median, intersection histograms of GoF or GoP Average, median, intersection histograms of GoF or GoP

1515

Figure 4: Three color images and their MPEG-7 histogram color distribution, depicted using a simplified color histogram. Based on the color distribution, the two left images would be recognized as more similar compared to the one on the right.

1616

Visual Texture DescriptorsVisual Texture Descriptors

Texture FeaturesTexture FeaturesVisual patterns (homogenous or non-Visual patterns (homogenous or non-

homogenous)homogenous)Multiple colors in imagesMultiple colors in imagesMultiple intensities in imagesMultiple intensities in imagesSurface structural information Surface structural information Figure 5Figure 5

1717

Figure 5: Examples of grayscale images with different textures. Using the MPEG-7 Visual texture descriptors, the two images on the bottom would be rated of similar texture, while less similar in texture compared to the two images on the top.

1818

Visual Texture DescriptorsVisual Texture Descriptors Homogenous texture descriptorHomogenous texture descriptor Figure 6Figure 6

Scale and orientation sensitive filtersScale and orientation sensitive filters Mean and SD of frequency coefficients (RT-FT)Mean and SD of frequency coefficients (RT-FT) Scale and rotation-invariant description and matchingScale and rotation-invariant description and matching 2D Gabor functions for filtering feature channels2D Gabor functions for filtering feature channels

Non homogenous texture descriptorNon homogenous texture descriptor (Edge Histogram) (Edge Histogram) Spatial distribution of edgesSpatial distribution of edges Division of image into 16 non overlapping blocks of equal Division of image into 16 non overlapping blocks of equal

sizesize Five edge categories: vertical, horizontal, 45 , 135 , and non Five edge categories: vertical, horizontal, 45 , 135 , and non

directional edge.directional edge. Rotation-sensitive and rotation-invariantRotation-sensitive and rotation-invariant Non uniform quantization using 3 bits, descriptor size of 240 Non uniform quantization using 3 bits, descriptor size of 240

bits (16x5x3)bits (16x5x3)

1919

Figure 6: Frequency layout for MPEG-7 Homogenous Texture Descriptor frequency extraction. Energy and energy deviation values are extracted from this frequency division into 30 channels.

2020

Visual Shape DescriptorsVisual Shape Descriptors

Provides a powerful visual clueProvides a powerful visual clue Invariant to scaling, rotation, and translationInvariant to scaling, rotation, and translation 2-D or 3-D in nature2-D or 3-D in nature

For 2-D there are two categoriesFor 2-D there are two categories Contour based which uses only boundary Contour based which uses only boundary

information of objectsinformation of objects Region-based which the entire shape regionRegion-based which the entire shape region

2121

Visual Shape DescriptorsVisual Shape Descriptors

3-D Shape Descriptor—Shape Spectrum Based on a shape spectrum concept Histogram of a shape index Measures local convexity of each local 3-D surface Histograms with 100 bins are used—each quantized

by 12 bits.

Region Based Shape Descriptor (Art) Figure 7 Uses all pixels constituting a shape within a frame Region-based moments invariant to transformations Coefficients of ART basis functions quantized

2222

Figure 7: Examples of various shapes that can be indexed using MPEG-7 Region-Based Shape Descriptor. Images contained in either of the sets (a)–(d) would be rated similar and dissimilar to the ones in the remaining sets. For example, images in set (a) would be identified being similar and dissimilar to the ones in set (b), (c), or (d).

2323

Contour based shape descriptor Contour based shape descriptor Figure 9Figure 9Curvature scale-space (CCS)Eccentricity and circularity valuesRobust to non-rigid motion partial occlusion of

the shape and perspective transformations2-D/3-D shape descriptor

Representation of 3-D objects using multiple 2-D snapshots

Types of Visual Shape DescriptorsTypes of Visual Shape Descriptors

2424

Figure 8: Examples of shapes that can be indexed using MPEG-7 Contour-Based Shape Descriptor.

2525

Motion Descriptors for VideoMotion Descriptors for Video Motion Activity DescriptorsMotion Activity Descriptors

Activity level and pace of motion in a sceneActivity level and pace of motion in a scene Motion activity intensity descriptorMotion activity intensity descriptor SD of motion vector magnitudeSD of motion vector magnitude SDs quantized into five activity levelsSDs quantized into five activity levels Optional FeaturesOptional Features

motion directionmotion direction spatial distribution of motion activityspatial distribution of motion activity Temporal distribution of motion activityTemporal distribution of motion activity

Camera Motion Descriptor Camera Motion Descriptor Figure 9Figure 9 Global motion parameters in timeGlobal motion parameters in time

zoom activityzoom activity translatory motiontranslatory motion

Motion similarity matching in particular time periodsMotion similarity matching in particular time periods

2626

Figure 9: Camera model for MPEG-7 Camera Motion Descriptor. Perspective projection to image plane p and camera motion parameters. The (virtual) camera is located in O.

2727

Warping Parameters Parametric motion descriptor Object description using 2-D parametric models

translations, rotations, scaling and combination of them planar perspective models quadratic models

Arbitrary objects, defined as regions (group of pixels) in the image over a specified time interval

Global sprite or mosaic Motion Trajectory Motion Trajectory

Description for independently moving objectsDescription for independently moving objects Object displacement over timeObject displacement over time

Motion Descriptors for VideoMotion Descriptors for Video

2828

Identify, filter and browse images using Identify, filter and browse images using visual contentvisual content

Specification to allow interoperability and Specification to allow interoperability and flexibilityflexibility

Other MPEG-7 standardsOther MPEG-7 standardsStorage, access and transmission of Storage, access and transmission of

descriptors and descriptors schemes in descriptors and descriptors schemes in system specificationsystem specification

ConclusionConclusion

2929

3030

Thank you