You are on page 1of 39

ChangeDetection

Preparedby: OluwafemiOpaleye

Objectives
Introduction WhatisChangeDetection? Preprocessing/Requirement ChangeDetectionTechniques ApplicationAreas PracticalExample FurtherReadings
2

04/07/2013

Introduction
Remote Sensing (RS) methods try to answer four basic q f questions:

HowmuchofWhatisWhere?
What:Type,CharacteristicandPropertiesofObject. e.g. g Water,Vegetation, g Landetc. HowMuch:determinebysimpleCounting, measuringAreacoveredorpercentageoftotalarea coverage. Where:Relatelocationsandareacoveredtoeithera standardmaportotheactuallocationonthe ground ground wheretheobjectoccurs. occurs 04/07/2013 Note:Where alsoreferstoamomentintime 3

WhatistheSHAPEandEXTENTof...? (Area,Boundaries (Area Boundaries,Lineaments Lineaments,...) ) ThisextendstheWHEREtobeacompletely GEOMETRICproblem. bl IdentificationandDelineationofBoundaries

04/07/2013

What atis st the eMIXo ofObjects? Thesurface Th f of fth theEarth E thi iscovered dby b objects bj t lik like Soil,Water,Grass,Trees,Houses,Roadsandsoon. Landuse/Landcover Classification

04/07/2013

HasitCHANGED?

CHANGEmayoccurwithprogressofTIME. Changemaybedetectedthroughcomparisonof observedstatesatdifferentmomentsintime. CHANGEDETECTION


04/07/2013 6

WhatisChangeDetection?
Changedetectionistheprocessofidentifying differencesinthestateofanobjectorphenomenon byobservingitatdifferenttimes. Itisthedetectionofclasstransitionbetweenapair ofcoregisteredimages images. Themaingoalistouseremotesensingtodetect CHANGEonalandscape(landuse andlandcover) overtime.
04/07/2013 7

Change detection algorithms analyze multiple images of the same scene taken at different times to identify y regions of change. Changes on the earth surface could be directly caused by natural forces, by the activities of animals and human i d d induced. Ti Timely l and d accurate change h d detection i of f Earths E h surface f features provides the foundation for a better understanding of the relationships and interactions between human and natural phenomena in order to better manage and use resources.
04/07/2013 8

It can be performed with raw remote sensing bands or thematic land cover maps classified from them. G Good d Change Ch D t ti research Detection h should h ld provide id the th following: areachange g rateofchange spatialdistributionofchangedtypes accuracyassessmentofchangedetectionresults

04/07/2013

Preprocessing/Requirement
GeometricCorrection Georeferencing precise coregistration betweenmultitemporal images RadiometricCorrection preciseradiometricand atmosphericcalibrationornormalizationbetween multitemporal images Region/Area / of fInterest samegeographic h location l Remotesensingsystemconsideration spatial, spectral,radiometricandtemporal
wheneverpossible,selectimagesacquiredfromthesame typeofsensors,withthesamespectralandspatial resolutions,andatthesameseasonaltimeframeinorderto minimize i i i unwanted t dvariances. i
04/07/2013 10

Freeofcloudsintheareaofanalysis Selecttimeperiods whatischangedetection period? SelectLandcover scheme theymustbeclassifiedin accordancewiththesameclassificationscheme scheme.


classesmustalsobedefinedidentically

Cl Classification ifi ti choose h classification l ifi ti algorithm l ith Choosechangedetectionmethod Changedetectionaccuracyassessment


04/07/2013 11

Majorstepsinvolvedinatypicalchangeanalysis processchange p g detectionp procedure

04/07/2013

12

ChangeDetectionTechniques
VisualAnalysis ImageDifferencing Imageratioing PostClassificationComparison Statisticalanalysis

04/07/2013

13

VisualAnalysis
Itisthefirstplacetostart Visuallycomparingmultiimages g gchanges g inmultiimages g isoften Manualdigitizing usedtobothidentifyandclassifychangebetween images g Elementsofimageinterpretationcombinedwiththe knowledgeoftheareaofstudyareoftenused used.

04/07/2013

14

DryingupofLakeFaguibine Mali
1974 2006

It covered area of about 590km2 Water level have fluctuated widely since the beginning of 1980 An extended period of reduced precipitation led to a complete drying of the lake
Source: Africa: Atlas of Our Changing Environment , UNEP
04/07/2013 15

DecliningWaterLevelsinLakeChad(19722007)
1972
A Area (12 (12,797sqkm) 797 k )

1987
Area (1,563sqkm)

2007
1987 Image g show that lake Chad reduced to about one-tenth of what it was in 1972 image. 2007 image show some improvement but the extent of the lake is still smaller to what it was 2-3 decades ago.
Area (1,753sqkm) (1 753sqkm)

Lake Chad, located at the junction of Niger, Nigeria, Chad and Cameroon, was once the sixth largest lake in the world. Persistent drought and increased agriculture irrigation have reduced the lakes extent

04/07/2013

16

Image g Differencing g
It requires selection of corresponding bands from two dates imageries of the same study area Uses software algorithm to identify and quantify the changes between two temporal images The difference image is created by subtracting the brightness g values of one image g from the other on a p per pixel basis. Unchanged g areas will have values at or nearer zero; ; while areas with significant change will be progressively positive or negative.
04/07/2013 17

Exampleofimagedifferencingprocedure

04/07/2013

18

Advantages
It is relatively y easy y to understand and to implement. p This method of analysis involves only subtraction with minimal human intervention. So long as the two images have been sampled to the same ground resolution and projected to the same coordinate system, the subtraction can be carried out very quickly. quickly The results of change detection are not subject to the inaccuracy inherent in classified land cover maps.
04/07/2013 19

Limitations
this method is limited in that it fails to reveal the nature of a detected change (e.g., the class from which a land cover has changed). identifythresholdvaluesofchangeandnochangein theresultingimages. direct use of raw spectral data in change analysis makes the detected change highly susceptible to radiometric variations caused by illumination conditions and seasonality.
04/07/2013 20

ImageRatioing
Similar to Image differencing conceptually and in its simplicity. This method uses one temporal image to divide image of another date. Values near to 1.0 indicate no change greater or less than 1.0 indicate changes g Values g Usually used for vegetation studies All other advantages and disadvantages of image differencing apply to image ratioing.
04/07/2013 21

Exampleofimageratioing procedure

04/07/2013

22

Post Classification Comparison


Mostpopularmethodofchangedetection Inpostclassificationcomparison,eachdateof rectifiedimageryisindependentlyclassifiedtofit commonlandtype. Landcover mapsareoverlaidandcomparedpixelby pixel p e bas basis. s Theresultisamapoflandtype change Thechangemapdisplayacreageofeachchange class
04/07/2013 23

Advantages
Manyclassificationalgorithmscanbedirectlyused. Itcanprovidedetailedmatrixofchangeinformation andaccuracyassessmentiseasy. Easytoquantifytheareaofchangeandrateof c a ge change Italsoattributechangese.g.

04/07/2013

24

Limitations
Classificationaccuracydirectlyinfluencesthe accuracyofchangedetection. Itistimeconsumingtocreateclassificationresults andaprofessionaloperatorisnecessary. Itisdifficultandexpensivetoobtainappropriate multi u temporal e po a ground g ou dreference. e e e ce

04/07/2013

25

SourcesofErrorinChangeDetection
Errorsindata imagequality Atmosphericerror Misregistrationbetweenmultipleimagedates Seasonalvariability Processingerror Radiometricerror duetosensordriftorage ErrorinClassification

04/07/2013

26

ApplicationAreas
landcover/landuse l d /l d changes h mappingurbangrowth rateofdeforestation urbansprawl desertification di t monitoring disaster it i agriculture coastalchange environmentalimpact p assessment
27

04/07/2013

PracticalExample: xample:
GeospatialAssessmentofAmanawa Forest Reserve,Sokoto State,Nigeria

04/07/2013

28

1996LandcoverMapofAmanawaForestReserve Area

04/07/2013

29

2008LandcoverMapofAmanawaForest ReserveArea

04/07/2013

30

Landcover Type

1996 A ( Area (sqkm) k )

1996 P Percentage t (%)

2008 A ( Area (sqkm) k )

2008 P Percentage t (%)

Farmland RockOutcrop BareSoil ForestReserve Dam Total

30.627 4.6449 3.537 2.0133 0.171 40.9932

74.71 11.33 8.63 4.91 0.42 100

30.772 4.0734 4.1517 1.89 0.1053 40.9932

75.07 9.94 10.12 4.61 0.26 100

04/07/2013

31

ChangeDetectionMapshowingtransitionof L d Landcovers (19962008)

04/07/2013

32

Landcover Type

Area( (sqkm) q )

Difference (sqkm) 1996 2008 0 145 0.145 0.572 0 572 0.615 0.123 0.066 0 066

Increase/Decline (%) 1996 2008 0 473 0.473 12.304 12 304 17.379 6.124 38.421 38 421
33

1996 Farmland 30 627 30.627 Rock Outcrop Outc op BareSoil Forest Reserve Dam
04/07/2013

2008 30 772 30.772 4 073 4.073 4.152 1.890 0 105 0.105

4 645 4.645 3.537 2.013 0 171 0.171

PredictionAnalysis y

04/07/2013

34

MarkovProbabilityofChangeinLandcover(1996 2008)
BareSoil Dam BareSoil Dam Farmland Forest 0.7646 0.2765
0.3212

Farmland Forest 0.1726 0.1098


0.6097

0.0382 0.6137 0.0680 0.0000 0.0000

0.0233 0.0000 0.0011 0.6676 0.0000

Rock Outcrop 0.0012 0.0000 0.0000 0.0000


0.6393

0.1849

0.1475 0.0231

RockOutcrop 0.3376

2018ProjectedLandcoverMapofAmanawaForest Area

AreaandPercentageof2018ProjectedLandcoverof AmanawaForestArea
Landcover Type Farmland Rock Outcrop Bare Soil Forest Reserve Dam Total Area (sqkm) 27 5877 27.5877 3.9555 7.6527 1.71 0.0873 40.9932 Percentage (%) 67 3 67.3 9.65 18.67 4.17 0.21 100

Reading gforfurtherinformation
J.R. J Je Jensen se ( (2005)Introductory 005) t oducto yDigital g ta Image age Processing,ARemotesensingperspective. 467492 R.R.Jensen,J.D.Gatrell andD.McLean (2007)GeoSpatialTechnologiesinUrban y,Practice, ,andPixels.145 EnvironmentsPolicy, 167

04/07/2013

38

ThankYouforListening

04/07/2013

39

You might also like