SlideShare uma empresa Scribd logo
1 de 26
Baixar para ler offline
Developing a classification framework for landcover
landuse change analysis in Chile
Dipl. Geoecologist Andreas Ch. Braun – Karlsruhe Institute of Technology – KIT


Institute of Photogrammetry and Remote Sensing - IPF




KIT – University of the State of Baden-Wuerttemberg and
National Research Center of the Helmholtz Association                      www.kit.edu
My background

      Andreas Ch. Braun – Diploma Geoecologist

      Works at the Institute of Photogrammetry and Remote Sensing
                Kernel-based (Vegetation) Classification
                    Support Vector Machines
                    Import Vector Machines
                    Relevance Vector Machines
                Feature Extraction Methods & Data Mining


      Received a special Ph. D. scholarship in 2010 from the german
      „Initiative for Excellence“

      For a case study on Deforestation and Forest Degradation in Chile



2    22.07.11        Dipl. Geoecologist Andreas Ch. Braun   Institute of Photogrammetry and Remote Sensing
The project on Deforestation in Chile

      Analyse impact of substitution of native forests with plantations (Pinus,
      Eucalyptus, Populus)
                Landscape fragmentation
                Habitat loss
                Biodiversity loss
      Approach:
      Biodiversity data (point data) in the field, interpolate via remote
      sensing/geoinformation on entire area (areal data)




3    22.07.11       Dipl. Geoecologist Andreas Ch. Braun   Institute of Photogrammetry and Remote Sensing
How can we get from here....




                                  Overall Accuracy 61,3%

4    22.07.11   Dipl. Geoecologist Andreas Ch. Braun       Institute of Photogrammetry and Remote Sensing
.... to here?




                                  Overall Accuracy 80,8% (+19,5)

5    22.07.11   Dipl. Geoecologist Andreas Ch. Braun       Institute of Photogrammetry and Remote Sensing
Review: Image Morphology




        Im. Matrix B                              Structuring Element S                     Im.Matrix B
     Erosion:         B⊖S :={z | Sz ⊆ B}                  →      All Pixels in S must be in foreground
     Dilatation:      B⊕S :={z | Sz ∩ B ≠ ∅} →                   Min. 1 Pixel in S must be in foreground
     Opening:         Erosion dann Dilatation
     Closing:         Dilatation dann Erosion




     Original                  Erosion                    Dilatation        Opening                       Closing
6    22.07.11      Dipl. Geoecologist Andreas Ch. Braun                        Institute of Photogrammetry and Remote Sensing
How can mathematical morphology help?




         Pinus radiata plantation




                                                        Populus nigra plantation
Nothofagus spec.
forest




 7    22.07.11   Dipl. Geoecologist Andreas Ch. Braun      Institute of Photogrammetry and Remote Sensing
How can mathematical morphology help?

      Toy-Example: Classification of plantations, forests, open soils




8    22.07.11   Dipl. Geoecologist Andreas Ch. Braun    Institute of Photogrammetry and Remote Sensing
How can mathematical morphology help?

      Toy-Example: Classification of plantations, forests, open soils




9    22.07.11   Dipl. Geoecologist Andreas Ch. Braun    Institute of Photogrammetry and Remote Sensing
How can mathematical morphology help?

       Toy-Example: Classification of plantations, forests, open soils




         Original                                          Opening               Closing




10    22.07.11      Dipl. Geoecologist Andreas Ch. Braun             Institute of Photogrammetry and Remote Sensing
How can mathematical morphology help?




       By using math. morphology, pixels are getting „more intelligent“. They
                 „know“ something about their neighbour pixels.

     Math. Morphology is one possibility of integrating the spatial context into a
                             spectral classification.

     „Mathematical morphology is a theory aiming to analyse the spatial
         relationships between pixels“ (Fauvel et al., 2008, p.3805)




11    22.07.11   Dipl. Geoecologist Andreas Ch. Braun     Institute of Photogrammetry and Remote Sensing
Morphological Attribute Profiles
      M. Dalla Mura, J. A. Benediktsson, B. Waske, L. Bruzzone (2010): „Morphological
      Attribute Profiles for the Analysis of Very High Resolution Images“. - IEEE
      Transactions on Geoscience and Remote Sensing, Vol. 48(10).

      Enhancements to the research on morphology in image classification by
      J.A.Benediktsson.

      Multilevel image analysis through opening, closing following these criteria:
             Area
             Moment of inertia
             Std. Deviation
             Diag. Of Bounding Box



      Not only one filter size but a vast range of different structuring elements.
      Graph-based approach increases computational performance.



12    22.07.11      Dipl. Geoecologist Andreas Ch. Braun             Institute of Photogrammetry and Remote Sensing
Graph-based approach

       Math. Morphology so far on binary images. How can grayscale images
       be used?
       Grayscale image is a stack of binary thresholds (e.g.. 8bit, [0,...,255])




     Intensity IKA                IKA > 80                  IKA > 120   IKA > 200                   IKA > 240

       Within this stack, a 256 level graph of connected components exits.




13    22.07.11       Dipl. Geoecologist Andreas Ch. Braun                  Institute of Photogrammetry and Remote Sensing
Morphological profile

       For these connected components (CC), certain criteria are checked
                 Area:        Is the area of a CC < the area of the structuring element ?
                 Inertia:     Is the extendedness of a CC < structuring element ?
                 Std. σ:      ...
                 Diag. BB:    ...


       If criteria are met, one image opening and one image closing is
       performed.

       Not only one structuring element is used, but an entire range →
       morphological profile.




14    22.07.11       Dipl. Geoecologist Andreas Ch. Braun             Institute of Photogrammetry and Remote Sensing
Morphological profile
      Afterwards, for classification we have:
                 One original image Im
                 Openings Opn, n=1,...,i,                  for different structuring elements
                 Closings Cln, n=1,...,i,                  for different structuring elements
      The morphological profile (MP) (Pesaresi, Benediktsson, 2000) is then:
                 MP={Cln, ...Im,...Opn}

     Cl3                  Cl2                     Cl1               Im             Op1              Op2                    Op3




      Instead of using only one channel and one MP, we can compute this on many
      channels, resulting in many Mps: extended morphological profile (EMP)
      (Benediktsson et al., 2005, Fauvel et al., 2008)
                 EMP={MPk1, MPk2, … , MPkm}



15    22.07.11           Dipl. Geoecologist Andreas Ch. Braun                             Institute of Photogrammetry and Remote Sensing
Additional features for classification

       For each channel of Landsat ETM+, we compute the features
                 Area:        2 per λ (Opening, Closing)
                 Inertia:     2 per λ
                 Std.:        2 per λ
                 Diag.BB:     2 per λ


       For 8 different λ

       8(features) * 8(channels) * 8(lambdas) = 512 new features for
       classification




16    22.07.11       Dipl. Geoecologist Andreas Ch. Braun   Institute of Photogrammetry and Remote Sensing
Classification of Landsat ETM+ image
                                                        3 Subsets

                                                        1: Forested area

                                                        2: Urban area

                                                        3: Agricultural area




17    22.07.11   Dipl. Geoecologist Andreas Ch. Braun    Institute of Photogrammetry and Remote Sensing
Subset 1: Forested area




                                   Overall Accuracy 61,3%

18    22.07.11   Dipl. Geoecologist Andreas Ch. Braun       Institute of Photogrammetry and Remote Sensing
Subset 1: Forested area




                                   Overall Accuracy 80,8% (+19,5)

19    22.07.11   Dipl. Geoecologist Andreas Ch. Braun       Institute of Photogrammetry and Remote Sensing
Subset 2: Urban area




                                   Overall Accuracy 75,5%

20    22.07.11   Dipl. Geoecologist Andreas Ch. Braun       Institute of Photogrammetry and Remote Sensing
Subset 2: Urban area




                                   Overall Accuracy 92,2% (+16,7)

21    22.07.11   Dipl. Geoecologist Andreas Ch. Braun       Institute of Photogrammetry and Remote Sensing
Subset 3: Agricultural area




                                   Overall Accuracy 62,2%

22    22.07.11   Dipl. Geoecologist Andreas Ch. Braun       Institute of Photogrammetry and Remote Sensing
Subset 3: Agricultural area




                                   Overall Accuracy 89,2% (+27,7)

23    22.07.11   Dipl. Geoecologist Andreas Ch. Braun       Institute of Photogrammetry and Remote Sensing
Conclusions

       Morphological Attribute Profiles are a very good, though implicit,
       method of integrating spatial context into spectrally motivated
       classification.

       Especially recommendable for classification of textured classed.

       Accuracy on three subsets in a image of Chile could be raised
       significantly.




24    22.07.11   Dipl. Geoecologist Andreas Ch. Braun    Institute of Photogrammetry and Remote Sensing
Challenges

       High dimensional feature space (>>500 features) can not be processed
       with standard methods (maximum likelihood).

       Specialized methods needed: kernel based:
                 Support vector machines
                 Import vector machines
                 Relevance vector machines


       Considerable programming effort.

       Computational expense requires high-perfomance PC (8-core
       processor with >120 GB Ram in our case)




25    22.07.11       Dipl. Geoecologist Andreas Ch. Braun   Institute of Photogrammetry and Remote Sensing
References
      M. Dalla Mura, J. A. Benediktsson, B. Waske, L. Bruzzone (2010): „Morphological Attribute
      Profiles for the Analysis of Very High Resolution Images“. - IEEE Transactions on Geoscience
      and Remote Sensing, Vol. 48(10).
      M. Fauvel, J.A. Benediktsson, J. Chanussot, J.R. Sveinsson (2008): „Spectral and Spatial
      Classification of Hyperspectral Data Using SVMs and Morphological Profiles“. - IEEE
      Transactions on Geoscience and Remote Sensing, Vol. 46(10).
      J.A. Benediktsson, J.A. Palmason, J.R. Sveinsson (2005): „Classification of Hyperspectral Data
      From Urban Areas Based on Extended Morphological Profiles“. - IEEE Transactions on
      Geoscience and Remote Sensing, Vol. 46(10).
      P. Soille, M. Pesaresi (2002): „Advances in mathematical morphology applied to geoscience and
      remote sensing“. - IEEE Transactions on Geoscience and Remote Sensing, Vol. 40(9).
      M. Pesaresi, J.A. Benediktsson (2000): „Image Segmentation based on the derivate of the
      morphological profile“.- In: Mathematical Morphology and Its Application to Image and Signal
      Processing, J. Goustsias, L. Vincent, D.S. Bloomberg, Eds. Norwell, MA: Kluwer, 2000.




26    22.07.11    Dipl. Geoecologist Andreas Ch. Braun                Institute of Photogrammetry and Remote Sensing

Mais conteúdo relacionado

Mais procurados

CLASSIFICATION AND COMPARISION OF REMOTE SENSING IMAGE USING SUPPORT VECTOR M...
CLASSIFICATION AND COMPARISION OF REMOTE SENSING IMAGE USING SUPPORT VECTOR M...CLASSIFICATION AND COMPARISION OF REMOTE SENSING IMAGE USING SUPPORT VECTOR M...
CLASSIFICATION AND COMPARISION OF REMOTE SENSING IMAGE USING SUPPORT VECTOR M...ADEIJ Journal
 
A SEMANTIC-BASED MULTILEVEL APPROACH TO CHANGE DETECTION IN VERY HIGH GEOMETR...
A SEMANTIC-BASED MULTILEVEL APPROACH TO CHANGE DETECTION IN VERY HIGH GEOMETR...A SEMANTIC-BASED MULTILEVEL APPROACH TO CHANGE DETECTION IN VERY HIGH GEOMETR...
A SEMANTIC-BASED MULTILEVEL APPROACH TO CHANGE DETECTION IN VERY HIGH GEOMETR...grssieee
 
Digital image processing and interpretation
Digital image processing and interpretationDigital image processing and interpretation
Digital image processing and interpretationP.K. Mani
 
A Comparative Study to Removal Salt & Pepper Noise from Satellite Image
A Comparative Study to Removal Salt & Pepper Noise from Satellite ImageA Comparative Study to Removal Salt & Pepper Noise from Satellite Image
A Comparative Study to Removal Salt & Pepper Noise from Satellite ImageIJCSIS Research Publications
 
satellite image processing
satellite image processingsatellite image processing
satellite image processingavhadlaxmikant
 
Improved single image dehazing by fusion
Improved single image dehazing by fusionImproved single image dehazing by fusion
Improved single image dehazing by fusioneSAT Publishing House
 
A Review over Different Blur Detection Techniques in Image Processing
A Review over Different Blur Detection Techniques in Image ProcessingA Review over Different Blur Detection Techniques in Image Processing
A Review over Different Blur Detection Techniques in Image Processingpaperpublications3
 
Basics of image processing & analysis
Basics of image processing & analysisBasics of image processing & analysis
Basics of image processing & analysisMohsin Siddique
 
RADAR Image Fusion Using Wavelet Transform
RADAR Image Fusion Using Wavelet TransformRADAR Image Fusion Using Wavelet Transform
RADAR Image Fusion Using Wavelet TransformINFOGAIN PUBLICATION
 
A Review on Haze Removal Techniques
A Review on Haze Removal TechniquesA Review on Haze Removal Techniques
A Review on Haze Removal TechniquesIRJET Journal
 
IRJET- Image Registration in GIS: A Survey
IRJET-  	  Image Registration in GIS: A SurveyIRJET-  	  Image Registration in GIS: A Survey
IRJET- Image Registration in GIS: A SurveyIRJET Journal
 
High Performance Computing for Satellite Image Processing and Analyzing – A ...
High Performance Computing for Satellite Image  Processing and Analyzing – A ...High Performance Computing for Satellite Image  Processing and Analyzing – A ...
High Performance Computing for Satellite Image Processing and Analyzing – A ...Editor IJCATR
 
Effective segmentation of sclera, iris and pupil in noisy eye images
Effective segmentation of sclera, iris and pupil in noisy eye imagesEffective segmentation of sclera, iris and pupil in noisy eye images
Effective segmentation of sclera, iris and pupil in noisy eye imagesTELKOMNIKA JOURNAL
 
2_ullo_presentation.pdf
2_ullo_presentation.pdf2_ullo_presentation.pdf
2_ullo_presentation.pdfgrssieee
 
76201940
7620194076201940
76201940IJRAT
 
Survey on Haze Removal Techniques
Survey on Haze Removal TechniquesSurvey on Haze Removal Techniques
Survey on Haze Removal TechniquesEditor IJMTER
 

Mais procurados (20)

CLASSIFICATION AND COMPARISION OF REMOTE SENSING IMAGE USING SUPPORT VECTOR M...
CLASSIFICATION AND COMPARISION OF REMOTE SENSING IMAGE USING SUPPORT VECTOR M...CLASSIFICATION AND COMPARISION OF REMOTE SENSING IMAGE USING SUPPORT VECTOR M...
CLASSIFICATION AND COMPARISION OF REMOTE SENSING IMAGE USING SUPPORT VECTOR M...
 
A SEMANTIC-BASED MULTILEVEL APPROACH TO CHANGE DETECTION IN VERY HIGH GEOMETR...
A SEMANTIC-BASED MULTILEVEL APPROACH TO CHANGE DETECTION IN VERY HIGH GEOMETR...A SEMANTIC-BASED MULTILEVEL APPROACH TO CHANGE DETECTION IN VERY HIGH GEOMETR...
A SEMANTIC-BASED MULTILEVEL APPROACH TO CHANGE DETECTION IN VERY HIGH GEOMETR...
 
Digital image processing and interpretation
Digital image processing and interpretationDigital image processing and interpretation
Digital image processing and interpretation
 
A Comparative Study to Removal Salt & Pepper Noise from Satellite Image
A Comparative Study to Removal Salt & Pepper Noise from Satellite ImageA Comparative Study to Removal Salt & Pepper Noise from Satellite Image
A Comparative Study to Removal Salt & Pepper Noise from Satellite Image
 
satellite image processing
satellite image processingsatellite image processing
satellite image processing
 
Improved single image dehazing by fusion
Improved single image dehazing by fusionImproved single image dehazing by fusion
Improved single image dehazing by fusion
 
A Review over Different Blur Detection Techniques in Image Processing
A Review over Different Blur Detection Techniques in Image ProcessingA Review over Different Blur Detection Techniques in Image Processing
A Review over Different Blur Detection Techniques in Image Processing
 
P180203105108
P180203105108P180203105108
P180203105108
 
Basics of image processing & analysis
Basics of image processing & analysisBasics of image processing & analysis
Basics of image processing & analysis
 
RADAR Image Fusion Using Wavelet Transform
RADAR Image Fusion Using Wavelet TransformRADAR Image Fusion Using Wavelet Transform
RADAR Image Fusion Using Wavelet Transform
 
A Review on Haze Removal Techniques
A Review on Haze Removal TechniquesA Review on Haze Removal Techniques
A Review on Haze Removal Techniques
 
IRJET- Image Registration in GIS: A Survey
IRJET-  	  Image Registration in GIS: A SurveyIRJET-  	  Image Registration in GIS: A Survey
IRJET- Image Registration in GIS: A Survey
 
High Performance Computing for Satellite Image Processing and Analyzing – A ...
High Performance Computing for Satellite Image  Processing and Analyzing – A ...High Performance Computing for Satellite Image  Processing and Analyzing – A ...
High Performance Computing for Satellite Image Processing and Analyzing – A ...
 
Effective segmentation of sclera, iris and pupil in noisy eye images
Effective segmentation of sclera, iris and pupil in noisy eye imagesEffective segmentation of sclera, iris and pupil in noisy eye images
Effective segmentation of sclera, iris and pupil in noisy eye images
 
Fd36957962
Fd36957962Fd36957962
Fd36957962
 
2_ullo_presentation.pdf
2_ullo_presentation.pdf2_ullo_presentation.pdf
2_ullo_presentation.pdf
 
Digital image processing
Digital image processingDigital image processing
Digital image processing
 
76201940
7620194076201940
76201940
 
Survey on Haze Removal Techniques
Survey on Haze Removal TechniquesSurvey on Haze Removal Techniques
Survey on Haze Removal Techniques
 
Publication
PublicationPublication
Publication
 

Semelhante a Developing a classification framework for landcover landuse change analysis in Chile

Geospatial Data Acquisition Using Unmanned Aerial Systems
Geospatial Data Acquisition Using Unmanned Aerial SystemsGeospatial Data Acquisition Using Unmanned Aerial Systems
Geospatial Data Acquisition Using Unmanned Aerial SystemsIEREK Press
 
Satellite image Processing Seminar Report
Satellite image Processing Seminar ReportSatellite image Processing Seminar Report
Satellite image Processing Seminar Reportalok ray
 
Panchromatic and Multispectral Remote Sensing Image Fusion Using Particle Swa...
Panchromatic and Multispectral Remote Sensing Image Fusion Using Particle Swa...Panchromatic and Multispectral Remote Sensing Image Fusion Using Particle Swa...
Panchromatic and Multispectral Remote Sensing Image Fusion Using Particle Swa...DR.P.S.JAGADEESH KUMAR
 
Earth Science and Remote Sensing Applications [Book]
Earth Science and Remote Sensing Applications [Book]Earth Science and Remote Sensing Applications [Book]
Earth Science and Remote Sensing Applications [Book]DR.P.S.JAGADEESH KUMAR
 
Tarımsal Toprak Haritalama'da Jeofizik Mühendisliği
Tarımsal Toprak Haritalama'da Jeofizik MühendisliğiTarımsal Toprak Haritalama'da Jeofizik Mühendisliği
Tarımsal Toprak Haritalama'da Jeofizik MühendisliğiAli Osman Öncel
 
Detection of Vegetation Coverage in Urban Agglomeration of Brussels by NDVI I...
Detection of Vegetation Coverage in Urban Agglomeration of Brussels by NDVI I...Detection of Vegetation Coverage in Urban Agglomeration of Brussels by NDVI I...
Detection of Vegetation Coverage in Urban Agglomeration of Brussels by NDVI I...Universität Salzburg
 
Kevin_Park_OSU_ Master_Project Report
Kevin_Park_OSU_ Master_Project ReportKevin_Park_OSU_ Master_Project Report
Kevin_Park_OSU_ Master_Project ReportKevin Park
 
Microscope_Telescope_p.pdf
Microscope_Telescope_p.pdfMicroscope_Telescope_p.pdf
Microscope_Telescope_p.pdfZero265663
 
Image Denoising Using Earth Mover's Distance and Local Histograms
Image Denoising Using Earth Mover's Distance and Local HistogramsImage Denoising Using Earth Mover's Distance and Local Histograms
Image Denoising Using Earth Mover's Distance and Local HistogramsCSCJournals
 
OBIA on Coastal Landform Based on Structure Tensor
OBIA on Coastal Landform Based on Structure Tensor OBIA on Coastal Landform Based on Structure Tensor
OBIA on Coastal Landform Based on Structure Tensor csandit
 
Supervised and unsupervised classification techniques for satellite imagery i...
Supervised and unsupervised classification techniques for satellite imagery i...Supervised and unsupervised classification techniques for satellite imagery i...
Supervised and unsupervised classification techniques for satellite imagery i...gaup_geo
 
Detection of Bridges using Different Types of High Resolution Satellite Images
Detection of Bridges using Different Types of High Resolution Satellite ImagesDetection of Bridges using Different Types of High Resolution Satellite Images
Detection of Bridges using Different Types of High Resolution Satellite Imagesidescitation
 
Quality assessment of data from CHRIS/PROBA
Quality assessment of data from CHRIS/PROBAQuality assessment of data from CHRIS/PROBA
Quality assessment of data from CHRIS/PROBAUniversität Salzburg
 
De-speckling of underwater ultrasound images
De-speckling of underwater ultrasound images De-speckling of underwater ultrasound images
De-speckling of underwater ultrasound images ajujohnkk
 
Computationally Efficient Methods for Sonar Image Denoising using Fractional ...
Computationally Efficient Methods for Sonar Image Denoising using Fractional ...Computationally Efficient Methods for Sonar Image Denoising using Fractional ...
Computationally Efficient Methods for Sonar Image Denoising using Fractional ...CSCJournals
 

Semelhante a Developing a classification framework for landcover landuse change analysis in Chile (20)

Geospatial Data Acquisition Using Unmanned Aerial Systems
Geospatial Data Acquisition Using Unmanned Aerial SystemsGeospatial Data Acquisition Using Unmanned Aerial Systems
Geospatial Data Acquisition Using Unmanned Aerial Systems
 
Satellite image Processing Seminar Report
Satellite image Processing Seminar ReportSatellite image Processing Seminar Report
Satellite image Processing Seminar Report
 
channel_mzhazbay.pdf
channel_mzhazbay.pdfchannel_mzhazbay.pdf
channel_mzhazbay.pdf
 
Panchromatic and Multispectral Remote Sensing Image Fusion Using Particle Swa...
Panchromatic and Multispectral Remote Sensing Image Fusion Using Particle Swa...Panchromatic and Multispectral Remote Sensing Image Fusion Using Particle Swa...
Panchromatic and Multispectral Remote Sensing Image Fusion Using Particle Swa...
 
Earth Science and Remote Sensing Applications [Book]
Earth Science and Remote Sensing Applications [Book]Earth Science and Remote Sensing Applications [Book]
Earth Science and Remote Sensing Applications [Book]
 
BREWSTER_K_poster
BREWSTER_K_posterBREWSTER_K_poster
BREWSTER_K_poster
 
Tarımsal Toprak Haritalama'da Jeofizik Mühendisliği
Tarımsal Toprak Haritalama'da Jeofizik MühendisliğiTarımsal Toprak Haritalama'da Jeofizik Mühendisliği
Tarımsal Toprak Haritalama'da Jeofizik Mühendisliği
 
Detection of Vegetation Coverage in Urban Agglomeration of Brussels by NDVI I...
Detection of Vegetation Coverage in Urban Agglomeration of Brussels by NDVI I...Detection of Vegetation Coverage in Urban Agglomeration of Brussels by NDVI I...
Detection of Vegetation Coverage in Urban Agglomeration of Brussels by NDVI I...
 
Kevin_Park_OSU_ Master_Project Report
Kevin_Park_OSU_ Master_Project ReportKevin_Park_OSU_ Master_Project Report
Kevin_Park_OSU_ Master_Project Report
 
F045033337
F045033337F045033337
F045033337
 
Microscope_Telescope_p.pdf
Microscope_Telescope_p.pdfMicroscope_Telescope_p.pdf
Microscope_Telescope_p.pdf
 
Image Denoising Using Earth Mover's Distance and Local Histograms
Image Denoising Using Earth Mover's Distance and Local HistogramsImage Denoising Using Earth Mover's Distance and Local Histograms
Image Denoising Using Earth Mover's Distance and Local Histograms
 
2
22
2
 
OBIA on Coastal Landform Based on Structure Tensor
OBIA on Coastal Landform Based on Structure Tensor OBIA on Coastal Landform Based on Structure Tensor
OBIA on Coastal Landform Based on Structure Tensor
 
Supervised and unsupervised classification techniques for satellite imagery i...
Supervised and unsupervised classification techniques for satellite imagery i...Supervised and unsupervised classification techniques for satellite imagery i...
Supervised and unsupervised classification techniques for satellite imagery i...
 
Detection of Bridges using Different Types of High Resolution Satellite Images
Detection of Bridges using Different Types of High Resolution Satellite ImagesDetection of Bridges using Different Types of High Resolution Satellite Images
Detection of Bridges using Different Types of High Resolution Satellite Images
 
Quality assessment of data from CHRIS/PROBA
Quality assessment of data from CHRIS/PROBAQuality assessment of data from CHRIS/PROBA
Quality assessment of data from CHRIS/PROBA
 
De-speckling of underwater ultrasound images
De-speckling of underwater ultrasound images De-speckling of underwater ultrasound images
De-speckling of underwater ultrasound images
 
Morpho
MorphoMorpho
Morpho
 
Computationally Efficient Methods for Sonar Image Denoising using Fractional ...
Computationally Efficient Methods for Sonar Image Denoising using Fractional ...Computationally Efficient Methods for Sonar Image Denoising using Fractional ...
Computationally Efficient Methods for Sonar Image Denoising using Fractional ...
 

Último

SOCIAL AND HISTORICAL CONTEXT - LFTVD.pptx
SOCIAL AND HISTORICAL CONTEXT - LFTVD.pptxSOCIAL AND HISTORICAL CONTEXT - LFTVD.pptx
SOCIAL AND HISTORICAL CONTEXT - LFTVD.pptxiammrhaywood
 
Key note speaker Neum_Admir Softic_ENG.pdf
Key note speaker Neum_Admir Softic_ENG.pdfKey note speaker Neum_Admir Softic_ENG.pdf
Key note speaker Neum_Admir Softic_ENG.pdfAdmir Softic
 
Interactive Powerpoint_How to Master effective communication
Interactive Powerpoint_How to Master effective communicationInteractive Powerpoint_How to Master effective communication
Interactive Powerpoint_How to Master effective communicationnomboosow
 
Software Engineering Methodologies (overview)
Software Engineering Methodologies (overview)Software Engineering Methodologies (overview)
Software Engineering Methodologies (overview)eniolaolutunde
 
Accessible design: Minimum effort, maximum impact
Accessible design: Minimum effort, maximum impactAccessible design: Minimum effort, maximum impact
Accessible design: Minimum effort, maximum impactdawncurless
 
BASLIQ CURRENT LOOKBOOK LOOKBOOK(1) (1).pdf
BASLIQ CURRENT LOOKBOOK  LOOKBOOK(1) (1).pdfBASLIQ CURRENT LOOKBOOK  LOOKBOOK(1) (1).pdf
BASLIQ CURRENT LOOKBOOK LOOKBOOK(1) (1).pdfSoniaTolstoy
 
Measures of Central Tendency: Mean, Median and Mode
Measures of Central Tendency: Mean, Median and ModeMeasures of Central Tendency: Mean, Median and Mode
Measures of Central Tendency: Mean, Median and ModeThiyagu K
 
Arihant handbook biology for class 11 .pdf
Arihant handbook biology for class 11 .pdfArihant handbook biology for class 11 .pdf
Arihant handbook biology for class 11 .pdfchloefrazer622
 
Q4-W6-Restating Informational Text Grade 3
Q4-W6-Restating Informational Text Grade 3Q4-W6-Restating Informational Text Grade 3
Q4-W6-Restating Informational Text Grade 3JemimahLaneBuaron
 
Student login on Anyboli platform.helpin
Student login on Anyboli platform.helpinStudent login on Anyboli platform.helpin
Student login on Anyboli platform.helpinRaunakKeshri1
 
APM Welcome, APM North West Network Conference, Synergies Across Sectors
APM Welcome, APM North West Network Conference, Synergies Across SectorsAPM Welcome, APM North West Network Conference, Synergies Across Sectors
APM Welcome, APM North West Network Conference, Synergies Across SectorsAssociation for Project Management
 
Kisan Call Centre - To harness potential of ICT in Agriculture by answer farm...
Kisan Call Centre - To harness potential of ICT in Agriculture by answer farm...Kisan Call Centre - To harness potential of ICT in Agriculture by answer farm...
Kisan Call Centre - To harness potential of ICT in Agriculture by answer farm...Krashi Coaching
 
9548086042 for call girls in Indira Nagar with room service
9548086042  for call girls in Indira Nagar  with room service9548086042  for call girls in Indira Nagar  with room service
9548086042 for call girls in Indira Nagar with room servicediscovermytutordmt
 
Measures of Dispersion and Variability: Range, QD, AD and SD
Measures of Dispersion and Variability: Range, QD, AD and SDMeasures of Dispersion and Variability: Range, QD, AD and SD
Measures of Dispersion and Variability: Range, QD, AD and SDThiyagu K
 
Unit-IV- Pharma. Marketing Channels.pptx
Unit-IV- Pharma. Marketing Channels.pptxUnit-IV- Pharma. Marketing Channels.pptx
Unit-IV- Pharma. Marketing Channels.pptxVishalSingh1417
 
Sports & Fitness Value Added Course FY..
Sports & Fitness Value Added Course FY..Sports & Fitness Value Added Course FY..
Sports & Fitness Value Added Course FY..Disha Kariya
 
Holdier Curriculum Vitae (April 2024).pdf
Holdier Curriculum Vitae (April 2024).pdfHoldier Curriculum Vitae (April 2024).pdf
Holdier Curriculum Vitae (April 2024).pdfagholdier
 
BAG TECHNIQUE Bag technique-a tool making use of public health bag through wh...
BAG TECHNIQUE Bag technique-a tool making use of public health bag through wh...BAG TECHNIQUE Bag technique-a tool making use of public health bag through wh...
BAG TECHNIQUE Bag technique-a tool making use of public health bag through wh...Sapna Thakur
 

Último (20)

SOCIAL AND HISTORICAL CONTEXT - LFTVD.pptx
SOCIAL AND HISTORICAL CONTEXT - LFTVD.pptxSOCIAL AND HISTORICAL CONTEXT - LFTVD.pptx
SOCIAL AND HISTORICAL CONTEXT - LFTVD.pptx
 
Key note speaker Neum_Admir Softic_ENG.pdf
Key note speaker Neum_Admir Softic_ENG.pdfKey note speaker Neum_Admir Softic_ENG.pdf
Key note speaker Neum_Admir Softic_ENG.pdf
 
Interactive Powerpoint_How to Master effective communication
Interactive Powerpoint_How to Master effective communicationInteractive Powerpoint_How to Master effective communication
Interactive Powerpoint_How to Master effective communication
 
Software Engineering Methodologies (overview)
Software Engineering Methodologies (overview)Software Engineering Methodologies (overview)
Software Engineering Methodologies (overview)
 
INDIA QUIZ 2024 RLAC DELHI UNIVERSITY.pptx
INDIA QUIZ 2024 RLAC DELHI UNIVERSITY.pptxINDIA QUIZ 2024 RLAC DELHI UNIVERSITY.pptx
INDIA QUIZ 2024 RLAC DELHI UNIVERSITY.pptx
 
Accessible design: Minimum effort, maximum impact
Accessible design: Minimum effort, maximum impactAccessible design: Minimum effort, maximum impact
Accessible design: Minimum effort, maximum impact
 
BASLIQ CURRENT LOOKBOOK LOOKBOOK(1) (1).pdf
BASLIQ CURRENT LOOKBOOK  LOOKBOOK(1) (1).pdfBASLIQ CURRENT LOOKBOOK  LOOKBOOK(1) (1).pdf
BASLIQ CURRENT LOOKBOOK LOOKBOOK(1) (1).pdf
 
Measures of Central Tendency: Mean, Median and Mode
Measures of Central Tendency: Mean, Median and ModeMeasures of Central Tendency: Mean, Median and Mode
Measures of Central Tendency: Mean, Median and Mode
 
Mattingly "AI & Prompt Design: Structured Data, Assistants, & RAG"
Mattingly "AI & Prompt Design: Structured Data, Assistants, & RAG"Mattingly "AI & Prompt Design: Structured Data, Assistants, & RAG"
Mattingly "AI & Prompt Design: Structured Data, Assistants, & RAG"
 
Arihant handbook biology for class 11 .pdf
Arihant handbook biology for class 11 .pdfArihant handbook biology for class 11 .pdf
Arihant handbook biology for class 11 .pdf
 
Q4-W6-Restating Informational Text Grade 3
Q4-W6-Restating Informational Text Grade 3Q4-W6-Restating Informational Text Grade 3
Q4-W6-Restating Informational Text Grade 3
 
Student login on Anyboli platform.helpin
Student login on Anyboli platform.helpinStudent login on Anyboli platform.helpin
Student login on Anyboli platform.helpin
 
APM Welcome, APM North West Network Conference, Synergies Across Sectors
APM Welcome, APM North West Network Conference, Synergies Across SectorsAPM Welcome, APM North West Network Conference, Synergies Across Sectors
APM Welcome, APM North West Network Conference, Synergies Across Sectors
 
Kisan Call Centre - To harness potential of ICT in Agriculture by answer farm...
Kisan Call Centre - To harness potential of ICT in Agriculture by answer farm...Kisan Call Centre - To harness potential of ICT in Agriculture by answer farm...
Kisan Call Centre - To harness potential of ICT in Agriculture by answer farm...
 
9548086042 for call girls in Indira Nagar with room service
9548086042  for call girls in Indira Nagar  with room service9548086042  for call girls in Indira Nagar  with room service
9548086042 for call girls in Indira Nagar with room service
 
Measures of Dispersion and Variability: Range, QD, AD and SD
Measures of Dispersion and Variability: Range, QD, AD and SDMeasures of Dispersion and Variability: Range, QD, AD and SD
Measures of Dispersion and Variability: Range, QD, AD and SD
 
Unit-IV- Pharma. Marketing Channels.pptx
Unit-IV- Pharma. Marketing Channels.pptxUnit-IV- Pharma. Marketing Channels.pptx
Unit-IV- Pharma. Marketing Channels.pptx
 
Sports & Fitness Value Added Course FY..
Sports & Fitness Value Added Course FY..Sports & Fitness Value Added Course FY..
Sports & Fitness Value Added Course FY..
 
Holdier Curriculum Vitae (April 2024).pdf
Holdier Curriculum Vitae (April 2024).pdfHoldier Curriculum Vitae (April 2024).pdf
Holdier Curriculum Vitae (April 2024).pdf
 
BAG TECHNIQUE Bag technique-a tool making use of public health bag through wh...
BAG TECHNIQUE Bag technique-a tool making use of public health bag through wh...BAG TECHNIQUE Bag technique-a tool making use of public health bag through wh...
BAG TECHNIQUE Bag technique-a tool making use of public health bag through wh...
 

Developing a classification framework for landcover landuse change analysis in Chile

  • 1. Developing a classification framework for landcover landuse change analysis in Chile Dipl. Geoecologist Andreas Ch. Braun – Karlsruhe Institute of Technology – KIT Institute of Photogrammetry and Remote Sensing - IPF KIT – University of the State of Baden-Wuerttemberg and National Research Center of the Helmholtz Association www.kit.edu
  • 2. My background Andreas Ch. Braun – Diploma Geoecologist Works at the Institute of Photogrammetry and Remote Sensing Kernel-based (Vegetation) Classification Support Vector Machines Import Vector Machines Relevance Vector Machines Feature Extraction Methods & Data Mining Received a special Ph. D. scholarship in 2010 from the german „Initiative for Excellence“ For a case study on Deforestation and Forest Degradation in Chile 2 22.07.11 Dipl. Geoecologist Andreas Ch. Braun Institute of Photogrammetry and Remote Sensing
  • 3. The project on Deforestation in Chile Analyse impact of substitution of native forests with plantations (Pinus, Eucalyptus, Populus) Landscape fragmentation Habitat loss Biodiversity loss Approach: Biodiversity data (point data) in the field, interpolate via remote sensing/geoinformation on entire area (areal data) 3 22.07.11 Dipl. Geoecologist Andreas Ch. Braun Institute of Photogrammetry and Remote Sensing
  • 4. How can we get from here.... Overall Accuracy 61,3% 4 22.07.11 Dipl. Geoecologist Andreas Ch. Braun Institute of Photogrammetry and Remote Sensing
  • 5. .... to here? Overall Accuracy 80,8% (+19,5) 5 22.07.11 Dipl. Geoecologist Andreas Ch. Braun Institute of Photogrammetry and Remote Sensing
  • 6. Review: Image Morphology Im. Matrix B Structuring Element S Im.Matrix B Erosion: B⊖S :={z | Sz ⊆ B} → All Pixels in S must be in foreground Dilatation: B⊕S :={z | Sz ∩ B ≠ ∅} → Min. 1 Pixel in S must be in foreground Opening: Erosion dann Dilatation Closing: Dilatation dann Erosion Original Erosion Dilatation Opening Closing 6 22.07.11 Dipl. Geoecologist Andreas Ch. Braun Institute of Photogrammetry and Remote Sensing
  • 7. How can mathematical morphology help? Pinus radiata plantation Populus nigra plantation Nothofagus spec. forest 7 22.07.11 Dipl. Geoecologist Andreas Ch. Braun Institute of Photogrammetry and Remote Sensing
  • 8. How can mathematical morphology help? Toy-Example: Classification of plantations, forests, open soils 8 22.07.11 Dipl. Geoecologist Andreas Ch. Braun Institute of Photogrammetry and Remote Sensing
  • 9. How can mathematical morphology help? Toy-Example: Classification of plantations, forests, open soils 9 22.07.11 Dipl. Geoecologist Andreas Ch. Braun Institute of Photogrammetry and Remote Sensing
  • 10. How can mathematical morphology help? Toy-Example: Classification of plantations, forests, open soils Original Opening Closing 10 22.07.11 Dipl. Geoecologist Andreas Ch. Braun Institute of Photogrammetry and Remote Sensing
  • 11. How can mathematical morphology help? By using math. morphology, pixels are getting „more intelligent“. They „know“ something about their neighbour pixels. Math. Morphology is one possibility of integrating the spatial context into a spectral classification. „Mathematical morphology is a theory aiming to analyse the spatial relationships between pixels“ (Fauvel et al., 2008, p.3805) 11 22.07.11 Dipl. Geoecologist Andreas Ch. Braun Institute of Photogrammetry and Remote Sensing
  • 12. Morphological Attribute Profiles M. Dalla Mura, J. A. Benediktsson, B. Waske, L. Bruzzone (2010): „Morphological Attribute Profiles for the Analysis of Very High Resolution Images“. - IEEE Transactions on Geoscience and Remote Sensing, Vol. 48(10). Enhancements to the research on morphology in image classification by J.A.Benediktsson. Multilevel image analysis through opening, closing following these criteria: Area Moment of inertia Std. Deviation Diag. Of Bounding Box Not only one filter size but a vast range of different structuring elements. Graph-based approach increases computational performance. 12 22.07.11 Dipl. Geoecologist Andreas Ch. Braun Institute of Photogrammetry and Remote Sensing
  • 13. Graph-based approach Math. Morphology so far on binary images. How can grayscale images be used? Grayscale image is a stack of binary thresholds (e.g.. 8bit, [0,...,255]) Intensity IKA IKA > 80 IKA > 120 IKA > 200 IKA > 240 Within this stack, a 256 level graph of connected components exits. 13 22.07.11 Dipl. Geoecologist Andreas Ch. Braun Institute of Photogrammetry and Remote Sensing
  • 14. Morphological profile For these connected components (CC), certain criteria are checked Area: Is the area of a CC < the area of the structuring element ? Inertia: Is the extendedness of a CC < structuring element ? Std. σ: ... Diag. BB: ... If criteria are met, one image opening and one image closing is performed. Not only one structuring element is used, but an entire range → morphological profile. 14 22.07.11 Dipl. Geoecologist Andreas Ch. Braun Institute of Photogrammetry and Remote Sensing
  • 15. Morphological profile Afterwards, for classification we have: One original image Im Openings Opn, n=1,...,i, for different structuring elements Closings Cln, n=1,...,i, for different structuring elements The morphological profile (MP) (Pesaresi, Benediktsson, 2000) is then: MP={Cln, ...Im,...Opn} Cl3 Cl2 Cl1 Im Op1 Op2 Op3 Instead of using only one channel and one MP, we can compute this on many channels, resulting in many Mps: extended morphological profile (EMP) (Benediktsson et al., 2005, Fauvel et al., 2008) EMP={MPk1, MPk2, … , MPkm} 15 22.07.11 Dipl. Geoecologist Andreas Ch. Braun Institute of Photogrammetry and Remote Sensing
  • 16. Additional features for classification For each channel of Landsat ETM+, we compute the features Area: 2 per λ (Opening, Closing) Inertia: 2 per λ Std.: 2 per λ Diag.BB: 2 per λ For 8 different λ 8(features) * 8(channels) * 8(lambdas) = 512 new features for classification 16 22.07.11 Dipl. Geoecologist Andreas Ch. Braun Institute of Photogrammetry and Remote Sensing
  • 17. Classification of Landsat ETM+ image 3 Subsets 1: Forested area 2: Urban area 3: Agricultural area 17 22.07.11 Dipl. Geoecologist Andreas Ch. Braun Institute of Photogrammetry and Remote Sensing
  • 18. Subset 1: Forested area Overall Accuracy 61,3% 18 22.07.11 Dipl. Geoecologist Andreas Ch. Braun Institute of Photogrammetry and Remote Sensing
  • 19. Subset 1: Forested area Overall Accuracy 80,8% (+19,5) 19 22.07.11 Dipl. Geoecologist Andreas Ch. Braun Institute of Photogrammetry and Remote Sensing
  • 20. Subset 2: Urban area Overall Accuracy 75,5% 20 22.07.11 Dipl. Geoecologist Andreas Ch. Braun Institute of Photogrammetry and Remote Sensing
  • 21. Subset 2: Urban area Overall Accuracy 92,2% (+16,7) 21 22.07.11 Dipl. Geoecologist Andreas Ch. Braun Institute of Photogrammetry and Remote Sensing
  • 22. Subset 3: Agricultural area Overall Accuracy 62,2% 22 22.07.11 Dipl. Geoecologist Andreas Ch. Braun Institute of Photogrammetry and Remote Sensing
  • 23. Subset 3: Agricultural area Overall Accuracy 89,2% (+27,7) 23 22.07.11 Dipl. Geoecologist Andreas Ch. Braun Institute of Photogrammetry and Remote Sensing
  • 24. Conclusions Morphological Attribute Profiles are a very good, though implicit, method of integrating spatial context into spectrally motivated classification. Especially recommendable for classification of textured classed. Accuracy on three subsets in a image of Chile could be raised significantly. 24 22.07.11 Dipl. Geoecologist Andreas Ch. Braun Institute of Photogrammetry and Remote Sensing
  • 25. Challenges High dimensional feature space (>>500 features) can not be processed with standard methods (maximum likelihood). Specialized methods needed: kernel based: Support vector machines Import vector machines Relevance vector machines Considerable programming effort. Computational expense requires high-perfomance PC (8-core processor with >120 GB Ram in our case) 25 22.07.11 Dipl. Geoecologist Andreas Ch. Braun Institute of Photogrammetry and Remote Sensing
  • 26. References M. Dalla Mura, J. A. Benediktsson, B. Waske, L. Bruzzone (2010): „Morphological Attribute Profiles for the Analysis of Very High Resolution Images“. - IEEE Transactions on Geoscience and Remote Sensing, Vol. 48(10). M. Fauvel, J.A. Benediktsson, J. Chanussot, J.R. Sveinsson (2008): „Spectral and Spatial Classification of Hyperspectral Data Using SVMs and Morphological Profiles“. - IEEE Transactions on Geoscience and Remote Sensing, Vol. 46(10). J.A. Benediktsson, J.A. Palmason, J.R. Sveinsson (2005): „Classification of Hyperspectral Data From Urban Areas Based on Extended Morphological Profiles“. - IEEE Transactions on Geoscience and Remote Sensing, Vol. 46(10). P. Soille, M. Pesaresi (2002): „Advances in mathematical morphology applied to geoscience and remote sensing“. - IEEE Transactions on Geoscience and Remote Sensing, Vol. 40(9). M. Pesaresi, J.A. Benediktsson (2000): „Image Segmentation based on the derivate of the morphological profile“.- In: Mathematical Morphology and Its Application to Image and Signal Processing, J. Goustsias, L. Vincent, D.S. Bloomberg, Eds. Norwell, MA: Kluwer, 2000. 26 22.07.11 Dipl. Geoecologist Andreas Ch. Braun Institute of Photogrammetry and Remote Sensing