moidex-the mobile tourist guide

Upload: chinmay451

Post on 08-Apr-2018

221 views

Category:

Documents


0 download

TRANSCRIPT

  • 8/7/2019 MoidEx-The Mobile Tourist Guide

    1/31

    Chinmay MohapatraENTC - B, 7th Semester

    Regn - 0601212451

  • 8/7/2019 MoidEx-The Mobile Tourist Guide

    2/31

  • 8/7/2019 MoidEx-The Mobile Tourist Guide

    3/31

  • 8/7/2019 MoidEx-The Mobile Tourist Guide

    4/31

    CONTENTS

    Introduction

    Moides Framework

    MoidEx Architecture MoidEx Location Detection

    Visual Feature Extraction

    mTourism Building Recognition System

    Mobile Configuration Requirements

    Performance Evaluation

    Conclusion

  • 8/7/2019 MoidEx-The Mobile Tourist Guide

    5/31

    INTRODUCTION

    According to World Tourism Org.

    In 2008, there were over 922

    million international touristarrivals, with a growth of 1.9% as

    compared to 2007.

    International tourism receipts

    grew to US $ 944 billion in 2008,

    corresponding to an increase inreal terms of 1.8% on 2007.

  • 8/7/2019 MoidEx-The Mobile Tourist Guide

    6/31

    Moides-Mobile Image Description

    MoidEx: The mTourism System

    The Moides gives the general framework for handset-

    based systems in order to extract features from images.

  • 8/7/2019 MoidEx-The Mobile Tourist Guide

    7/31

    MoidEx Architecture

    TAJ.MPEG

    TEXT

    MPEG-7

    CBIR

    Mpeg7 reference

    database

    CLC

  • 8/7/2019 MoidEx-The Mobile Tourist Guide

    8/31

    MPEG-7 is an ISO/IEC standard developed by

    Moving Pictures Experts Group (MPEG).

    MPEG-7 helps in describing multimediacontent in a standardized way and thus

    enables easier archiving, accessing, locating,

    navigating, searching, and managing of

    audiovisual information.

    MPEG-7

  • 8/7/2019 MoidEx-The Mobile Tourist Guide

    9/31

    Location Detection-CELL ID

    GSM CELL

    Mobile Network Code

    (MNC)

    Mobile Country Code

    (MCC)

    Cell ID (CID)

    Location Area Code

    (LAC)

  • 8/7/2019 MoidEx-The Mobile Tourist Guide

    10/31

    Location Detection-CELL ID

    URBAN

    RURAL AREA

    RANGE ADVANTAGES

    1. High availability.

    2. Low cost.

    3. Doesnt require any

    modification in handset of

    cellular network.

  • 8/7/2019 MoidEx-The Mobile Tourist Guide

    11/31

    MoidEx uses Cell ID approach.

    The general Idea is to determine the current

    cell and using its centre as the approximated

    location as its handset.

    MoidEx Location Detection

  • 8/7/2019 MoidEx-The Mobile Tourist Guide

    12/31

    < Cr e a t i o n L o c a t i o n>

    IMEI = 358361007597812 ,

    MCC = 262 ,

    MNC = 01 ,

    LAC = 33028 ,

    CID = 63857 ,

    SIGNAL = 94 ,

    BAR = 3

    Eg. of a creation location descriptor, created outside a street in London.

    MoidEx Location Detection

  • 8/7/2019 MoidEx-The Mobile Tourist Guide

    13/31

    VISUAL FEATUREEXTRACTION

    CLC ALGORITHM

    A Consistent Line Cluster (CLC) is a set of lines that

    are homogeneous in terms of some line features.

    Color-CLC: The lines have the same color feature.

    Orientation-CLC: The lines are parallel to each other.

    Spatially-CLC: The lines are in close proximity to each

    other.

  • 8/7/2019 MoidEx-The Mobile Tourist Guide

    14/31

  • 8/7/2019 MoidEx-The Mobile Tourist Guide

    15/31

    To reduce the complexity, classify each pixel of the image as

    one of several dominant colors.

    Color feature of lines: Color Pair (c1,c2).

  • 8/7/2019 MoidEx-The Mobile Tourist Guide

    16/31

    (contd.)

    Lines are clustered according to their color pair(s).

  • 8/7/2019 MoidEx-The Mobile Tourist Guide

    17/31

    The lines in an Orientation-CLC are parallel to each other in

    the 3D world; the parallel lines of an object in a 2D image can

    be:

    Still parallel.

    Converging to a vanishing point.

  • 8/7/2019 MoidEx-The Mobile Tourist Guide

    18/31

    (contd.)

    Two steps:

    Construct roughly-orientation-CLC;

    Filter line clusters using perspective information.

  • 8/7/2019 MoidEx-The Mobile Tourist Guide

    19/31

    Vertical position clustering

    Horizontal position clustering

  • 8/7/2019 MoidEx-The Mobile Tourist Guide

    20/31

    Two types ofbuildings Two criteria:

    Inter-relationship criterion

  • 8/7/2019 MoidEx-The Mobile Tourist Guide

    21/31

    (contd.)

    Intra-relationship criterion

  • 8/7/2019 MoidEx-The Mobile Tourist Guide

    22/31

    MAPPING CLC INTO MPEG-7

    Given a consistent line cluster cli with lines:

    l1= ((x11, y11)(x12, y12)), , ln = ((xn1, yn1)(xn2, yn2))

    the proposed mapping of this cluster to the MPEG-7 RegionLocatorType

    descriptor is shown in Listening 2.

    1 2 3 4 x11 y11 x12 y125 . . .

    6 xn1 yn1 xn2 yn27 8 9

    Mapping of line clusters to polygons

  • 8/7/2019 MoidEx-The Mobile Tourist Guide

    23/31

    MAPPING CLC INTO MPEG-7

    Moides

    CLC

    Algorithm

    SELECT

    Moides

    CANCEL

    ProcessingProcessing

    Taj.mpeg

    BACK

    1 2 3

  • 8/7/2019 MoidEx-The Mobile Tourist Guide

    24/31

    mTourism Building Recognition System

    The mTourism building recognition system has

    been developed as application library on topof MPEG-7 MMDB system.

    The MPEG-7 MMDB system provides means

    for storing, indexing, querying MPEG-7

    instance documents.

  • 8/7/2019 MoidEx-The Mobile Tourist Guide

    25/31

    KNN-retrieval for mTourism Lib within MPEG-7 MMDB

  • 8/7/2019 MoidEx-The Mobile Tourist Guide

    26/31

    Mobile Configuration Requirements

    1. Camera Phone (at least 2 megapixel).

    2. Connectivity: Bluetooth.

    GPRS.

    3. Java Micro Edition (J2ME).

  • 8/7/2019 MoidEx-The Mobile Tourist Guide

    27/31

    Performance EvaluationPerformance Evaluation

    97 well-patterned buildings : 97/97

    44 not-well-patterned buildings : 42/44

    16 non-patterned non-buildings : 15/16

    25 patterned non-buildings : 0/25

  • 8/7/2019 MoidEx-The Mobile Tourist Guide

    28/31

  • 8/7/2019 MoidEx-The Mobile Tourist Guide

    29/31

    Conclusion

    MoidEx enhanced the original framework by

    the means of additional visual features

    location detection and an algorithm targeting

    on building recognition.

    The evaluation showed a recognition rate up

    to 90% by combining location and visual

    features.

    The average error was 5.8%.

  • 8/7/2019 MoidEx-The Mobile Tourist Guide

    30/31

    REFERENCES

    MoidEx: Location-based mTourism system on mobile devices.M.Dller, G. Kckerandl, S. Jans, L.Limam. DansInternational Conference on Multimedia Computing andSystems, ICMCS'2009, IEEE ed. Ouarzazate, Maroc. 2009.

    Moides - Mobile Image Description

    Mario Doller and Harald Kosch and Josef Hammer jun.Proceedings of I-KNOW 06 Graz, Austria, September 6 - 8, 2006 .

    Consistent line clusters for building recognition in CBIR.

    Yi Li and Linda G. Shapiro. In Proceedings of the 16thInternational Conference on Pattern Recognition (ICPR),pages 952956, Washington, DC, USA, 2002. IEEEComputer Society.

  • 8/7/2019 MoidEx-The Mobile Tourist Guide

    31/31