SlideShare a Scribd company logo
OSGEO-India: FOSS4G 2012- First National Conference "Open Source Geospatial Resources
     to Spearhead Development and Growth” 25-27th October 2012, @ IIIT Hyderabad




     Object Based Image Analysis
          Tools for Opticks
                    Mohit Kumar, KS Rajan, Dustan Adkins




                                http://osgeo.in/foss4g2012                              1
OPTICKS ?
    •   Opticks is an open source, remote sensing application that supports imagery,
        video (motion imagery), Synthetic Aperture Radar (SAR), multi-spectral, hyper-
        spectral, and other types of remote sensing data.
    •   Opticks can also be used as a remote sensing software development framework.
        Developers can extend Opticks functionality using its plug-in architecture and
        public application programming interface
    •   http://opticks.org


                     Why object-based?
•   Object based approach is better than conventional per-pixel analysis as it deals
    with considerably reduced number of units. This approach is not that much
    sensitive to noise and hence is spatially consistent.




                                     http://osgeo.in/foss4g2012                          2
Workflow




http://osgeo.in/foss4g2012   3
Image Segmentation (Meanshift)
                                                            Input image
                                                           (CIELAB colour          5 Dimensional
 Input image(RGB)                                                                  feature space
                                                               space)


                                             sing           Modes(local
                                      o lla p
                               space c ect.                                          Clustering
                         ature      ob j
                                                             maximas)
                    in fe form an
              i n ts o d e
           Po
                    em
            to o n


 Objects formed
                                                            Pruning (spatial     Pruning (Spectral
(backtracking the
                                                              Bandwidth)            Bandwidth)
     modes)




                                                                                 Pruning ( Minimum
                                                                                    region area)

                                                    http://osgeo.in/foss4g2012                       4
Object attribution
• Calculating textural, geometric and spectral features for the objects
  made in the Segmentation step in a feature vector.
• Area, Perimeter, Roundness, Compactness, Centroid, Contrast,
  Coarseness, Direction, Roughness, Mean red, Mean green, Mean
  blue, std. deviation Red, std. deviation Green, std. deviation Blue.


     Segmented Image                            For every object in the image



                                                Initialize a vector having all 16
                                                             features


                                               Calculate value for every feature
                                                    and save in the vector.

                            http://osgeo.in/foss4g2012                              5
Classification
• Mahalanobis Distance
• Di,j2 = (x-µj)`S-1(x-µj)
• The class which has the least Mahalanobis
  distance to the object i is the class of that object.

                      Vectorization
 • Creates vector polygons for all connected regions of pixels in the
   object image sharing a common pixel value.
 • Polygon features are created on the output layer, with polygon
   geometries representing the polygons.
 • The class which has the least Mahalanobis distance to the object i is
   the class of that object.


                          http://osgeo.in/foss4g2012                 6
The Input Orbview3 (4m) data of a part of delhi (500X500)                    The output of the objects with area less than 100.




                                                                               The shapefile(.shp) displaying the objects
Output of object having area 100-200 and classified as building.


                                                      http://osgeo.in/foss4g2012                                                   7
Performance Analysis

 Image Size         Number of          Running time
                     objects              (sec)
  512 X 512           153                  3.1

 1024 X 1024            435                  24.55


                Table 1 : Image segmentation


Image Size       Number of objects       Running time
                                            (sec)
 256 X 256               49                 0.249
 512 X 512               193                    1.133
1024 X1024               752                    5.359
2048 X 2048             2965                    31.60
4096 X 4096            11922                    286.59

              Table 2 : Object Attribution



                   http://osgeo.in/foss4g2012            8
Source Code
• https://github.com/mohitkharb/Opticks_GSO
  C2012
• http://opticks.org/confluence/display/~mohit
  kharb/Workflow+of+the+pluggin
• http://code.google.com/soc/




                  http://osgeo.in/foss4g2012     9
Any Questions?




  http://osgeo.in/foss4g2012   10

More Related Content

What's hot (20)

PPTX
Chap. 10 computational photography
duckleek
 
PPTX
Edge Detection
Jakir Hossain
 
PPTX
Edge detection using evolutionary algorithms new
Priyanka Sharma
 
PPTX
Computer vision
Antonio Radesca
 
PPTX
[DL輪読会]ClearGrasp
Deep Learning JP
 
PPTX
Comparison between Blur Transfer and Blur Re-Generation in Depth Image Based ...
Norishige Fukushima
 
PPT
CS 354 Acceleration Structures
Mark Kilgard
 
PDF
Thesis presentation
Caroline Pacheco do Espírito Silva
 
PPTX
Yonsei Data Science Lab - Computer Vision
DataScienceLab
 
PDF
Feature based ghost removal in high dynamic range imaging
ijcga
 
PPTX
Canny edge detection
ahmedkhaledfayez
 
PPTX
Fuzzy Logic Based Edge Detection
Dawn Raider Gupta
 
PPTX
Final Review
Afzal Meeran
 
PPT
Real-time Shadowing Techniques: Shadow Volumes
Mark Kilgard
 
PDF
image segmentation
arpanmankar
 
PPTX
Shiny Pixels and Beyond: Real-Time Raytracing at SEED
Electronic Arts / DICE
 
PPTX
Background subtraction
Shashank Dhariwal
 
DOC
Seminar report on edge detection of video using matlab code
Bhushan Deore
 
PPT
03 digital image fundamentals DIP
babak danyal
 
PDF
ALGORITHM AND TECHNIQUE ON VARIOUS EDGE DETECTION: A SURVEY
sipij
 
Chap. 10 computational photography
duckleek
 
Edge Detection
Jakir Hossain
 
Edge detection using evolutionary algorithms new
Priyanka Sharma
 
Computer vision
Antonio Radesca
 
[DL輪読会]ClearGrasp
Deep Learning JP
 
Comparison between Blur Transfer and Blur Re-Generation in Depth Image Based ...
Norishige Fukushima
 
CS 354 Acceleration Structures
Mark Kilgard
 
Yonsei Data Science Lab - Computer Vision
DataScienceLab
 
Feature based ghost removal in high dynamic range imaging
ijcga
 
Canny edge detection
ahmedkhaledfayez
 
Fuzzy Logic Based Edge Detection
Dawn Raider Gupta
 
Final Review
Afzal Meeran
 
Real-time Shadowing Techniques: Shadow Volumes
Mark Kilgard
 
image segmentation
arpanmankar
 
Shiny Pixels and Beyond: Real-Time Raytracing at SEED
Electronic Arts / DICE
 
Background subtraction
Shashank Dhariwal
 
Seminar report on edge detection of video using matlab code
Bhushan Deore
 
03 digital image fundamentals DIP
babak danyal
 
ALGORITHM AND TECHNIQUE ON VARIOUS EDGE DETECTION: A SURVEY
sipij
 

Viewers also liked (18)

PDF
Revista Adopta'm
Animals del Maresme
 
PPT
Hivern 2012 - 2013
Animals del Maresme
 
PPTX
ისტ ისგამოყენება
atojanraisa
 
PDF
Revista juny 2013
Animals del Maresme
 
PPT
Protagonistes de la 19a passejada
Animals del Maresme
 
PPT
Animaladda 2012
Animals del Maresme
 
PPTX
2014年 忘年会 歳出
Tatsuya Suzuki
 
PDF
Revista febrer 2013
Animals del Maresme
 
PPTX
ისტ ისგამოყენება
atojanraisa
 
PDF
ACAAD Maresme - Balanç 2013-2014
Animals del Maresme
 
PPS
Presentació Animaladda 2014
Animals del Maresme
 
PDF
Inversions i activitats 2013-2015
Animals del Maresme
 
PDF
โครงการระบบดูแล ปี 55
somchaitumdee50
 
PDF
โครงการส่งเสริมคุณธรรมจริยธรรม ค่านิยมที่พึงประสงค์.54
somchaitumdee50
 
PDF
โครงการระบบดูแล ปี 55
somchaitumdee50
 
PDF
คำสั่งของโรงเรียอาจารย์ที่ปรึกษา.Doc ปี 2555
somchaitumdee50
 
PPT
Els protagonistes de la 21a passejada
Animals del Maresme
 
PDF
แผน พล ม. 1 ภาค 1 ปี 54 ปิงปอง
somchaitumdee50
 
Revista Adopta'm
Animals del Maresme
 
Hivern 2012 - 2013
Animals del Maresme
 
ისტ ისგამოყენება
atojanraisa
 
Revista juny 2013
Animals del Maresme
 
Protagonistes de la 19a passejada
Animals del Maresme
 
Animaladda 2012
Animals del Maresme
 
2014年 忘年会 歳出
Tatsuya Suzuki
 
Revista febrer 2013
Animals del Maresme
 
ისტ ისგამოყენება
atojanraisa
 
ACAAD Maresme - Balanç 2013-2014
Animals del Maresme
 
Presentació Animaladda 2014
Animals del Maresme
 
Inversions i activitats 2013-2015
Animals del Maresme
 
โครงการระบบดูแล ปี 55
somchaitumdee50
 
โครงการส่งเสริมคุณธรรมจริยธรรม ค่านิยมที่พึงประสงค์.54
somchaitumdee50
 
โครงการระบบดูแล ปี 55
somchaitumdee50
 
คำสั่งของโรงเรียอาจารย์ที่ปรึกษา.Doc ปี 2555
somchaitumdee50
 
Els protagonistes de la 21a passejada
Animals del Maresme
 
แผน พล ม. 1 ภาค 1 ปี 54 ปิงปอง
somchaitumdee50
 
Ad

Similar to Object based image analysis tools for opticks (20)

PDF
Modern features-part-2-descriptors
zukun
 
PDF
IAP09 CUDA@MIT 6.963 - Lecture 01: High-Throughput Scientific Computing (Hans...
npinto
 
PDF
MIT 6.870 - Template Matching and Histograms (Nicolas Pinto, MIT)
npinto
 
PDF
Mit6870 template matching and histograms
zukun
 
ZIP
ICPRAM 2012
Isa Restrepo
 
PDF
Parallel Distributed Image Stacking and Mosaicing with Hadoop__HadoopSummit2010
Yahoo Developer Network
 
PPTX
3rd Seminar
ChristinaSaweres
 
PPTX
3rd Seminar
ChristinaSaweres
 
PDF
IRJET-Comparison of SIFT & SURF Corner Detector as Features and other Machine...
IRJET Journal
 
PDF
igarss.pdf
grssieee
 
PDF
C42011318
IJERA Editor
 
PDF
What is Object-Based Analysis
Esri
 
PDF
Illumination-robust Recognition and Inspection in Carbide Insert Production
IDES Editor
 
PDF
Lecture 02 internet video search
zukun
 
PDF
Modern features-part-0-intro
zukun
 
PPT
SIFT.ppt
usatwikaa
 
PDF
Pc Seminar Jordi
Universitat de Barcelona
 
PDF
Fcv core szeliski_zisserman
zukun
 
PDF
Advanced computer vision transfroemasfgmblzfbmlzamfgvDLMV.pdf
SamruddhiChillure1
 
PDF
Focus set based reconstruction of micro-objects
Jan Wedekind
 
Modern features-part-2-descriptors
zukun
 
IAP09 CUDA@MIT 6.963 - Lecture 01: High-Throughput Scientific Computing (Hans...
npinto
 
MIT 6.870 - Template Matching and Histograms (Nicolas Pinto, MIT)
npinto
 
Mit6870 template matching and histograms
zukun
 
ICPRAM 2012
Isa Restrepo
 
Parallel Distributed Image Stacking and Mosaicing with Hadoop__HadoopSummit2010
Yahoo Developer Network
 
3rd Seminar
ChristinaSaweres
 
3rd Seminar
ChristinaSaweres
 
IRJET-Comparison of SIFT & SURF Corner Detector as Features and other Machine...
IRJET Journal
 
igarss.pdf
grssieee
 
C42011318
IJERA Editor
 
What is Object-Based Analysis
Esri
 
Illumination-robust Recognition and Inspection in Carbide Insert Production
IDES Editor
 
Lecture 02 internet video search
zukun
 
Modern features-part-0-intro
zukun
 
SIFT.ppt
usatwikaa
 
Pc Seminar Jordi
Universitat de Barcelona
 
Fcv core szeliski_zisserman
zukun
 
Advanced computer vision transfroemasfgmblzfbmlzamfgvDLMV.pdf
SamruddhiChillure1
 
Focus set based reconstruction of micro-objects
Jan Wedekind
 
Ad

Recently uploaded (20)

PPT
DRUGS USED IN THERAPY OF SHOCK, Shock Therapy, Treatment or management of shock
Rajshri Ghogare
 
PPTX
Virus sequence retrieval from NCBI database
yamunaK13
 
PPTX
How to Close Subscription in Odoo 18 - Odoo Slides
Celine George
 
PPTX
I INCLUDED THIS TOPIC IS INTELLIGENCE DEFINITION, MEANING, INDIVIDUAL DIFFERE...
parmarjuli1412
 
PPTX
How to Track Skills & Contracts Using Odoo 18 Employee
Celine George
 
PPTX
Cleaning Validation Ppt Pharmaceutical validation
Ms. Ashatai Patil
 
PPTX
Basics and rules of probability with real-life uses
ravatkaran694
 
PPTX
Sonnet 130_ My Mistress’ Eyes Are Nothing Like the Sun By William Shakespear...
DhatriParmar
 
PPTX
PROTIEN ENERGY MALNUTRITION: NURSING MANAGEMENT.pptx
PRADEEP ABOTHU
 
PDF
The-Invisible-Living-World-Beyond-Our-Naked-Eye chapter 2.pdf/8th science cur...
Sandeep Swamy
 
PPTX
Artificial Intelligence in Gastroentrology: Advancements and Future Presprec...
AyanHossain
 
PDF
Antianginal agents, Definition, Classification, MOA.pdf
Prerana Jadhav
 
PPTX
ENGLISH 8 WEEK 3 Q1 - Analyzing the linguistic, historical, andor biographica...
OliverOllet
 
PPTX
Gupta Art & Architecture Temple and Sculptures.pptx
Virag Sontakke
 
PPTX
Translation_ Definition, Scope & Historical Development.pptx
DhatriParmar
 
PDF
My Thoughts On Q&A- A Novel By Vikas Swarup
Niharika
 
PPTX
Electrophysiology_of_Heart. Electrophysiology studies in Cardiovascular syste...
Rajshri Ghogare
 
PPTX
Top 10 AI Tools, Like ChatGPT. You Must Learn In 2025
Digilearnings
 
PPTX
K-Circle-Weekly-Quiz12121212-May2025.pptx
Pankaj Rodey
 
PDF
John Keats introduction and list of his important works
vatsalacpr
 
DRUGS USED IN THERAPY OF SHOCK, Shock Therapy, Treatment or management of shock
Rajshri Ghogare
 
Virus sequence retrieval from NCBI database
yamunaK13
 
How to Close Subscription in Odoo 18 - Odoo Slides
Celine George
 
I INCLUDED THIS TOPIC IS INTELLIGENCE DEFINITION, MEANING, INDIVIDUAL DIFFERE...
parmarjuli1412
 
How to Track Skills & Contracts Using Odoo 18 Employee
Celine George
 
Cleaning Validation Ppt Pharmaceutical validation
Ms. Ashatai Patil
 
Basics and rules of probability with real-life uses
ravatkaran694
 
Sonnet 130_ My Mistress’ Eyes Are Nothing Like the Sun By William Shakespear...
DhatriParmar
 
PROTIEN ENERGY MALNUTRITION: NURSING MANAGEMENT.pptx
PRADEEP ABOTHU
 
The-Invisible-Living-World-Beyond-Our-Naked-Eye chapter 2.pdf/8th science cur...
Sandeep Swamy
 
Artificial Intelligence in Gastroentrology: Advancements and Future Presprec...
AyanHossain
 
Antianginal agents, Definition, Classification, MOA.pdf
Prerana Jadhav
 
ENGLISH 8 WEEK 3 Q1 - Analyzing the linguistic, historical, andor biographica...
OliverOllet
 
Gupta Art & Architecture Temple and Sculptures.pptx
Virag Sontakke
 
Translation_ Definition, Scope & Historical Development.pptx
DhatriParmar
 
My Thoughts On Q&A- A Novel By Vikas Swarup
Niharika
 
Electrophysiology_of_Heart. Electrophysiology studies in Cardiovascular syste...
Rajshri Ghogare
 
Top 10 AI Tools, Like ChatGPT. You Must Learn In 2025
Digilearnings
 
K-Circle-Weekly-Quiz12121212-May2025.pptx
Pankaj Rodey
 
John Keats introduction and list of his important works
vatsalacpr
 

Object based image analysis tools for opticks

  • 1. OSGEO-India: FOSS4G 2012- First National Conference "Open Source Geospatial Resources to Spearhead Development and Growth” 25-27th October 2012, @ IIIT Hyderabad Object Based Image Analysis Tools for Opticks Mohit Kumar, KS Rajan, Dustan Adkins http://osgeo.in/foss4g2012 1
  • 2. OPTICKS ? • Opticks is an open source, remote sensing application that supports imagery, video (motion imagery), Synthetic Aperture Radar (SAR), multi-spectral, hyper- spectral, and other types of remote sensing data. • Opticks can also be used as a remote sensing software development framework. Developers can extend Opticks functionality using its plug-in architecture and public application programming interface • http://opticks.org Why object-based? • Object based approach is better than conventional per-pixel analysis as it deals with considerably reduced number of units. This approach is not that much sensitive to noise and hence is spatially consistent. http://osgeo.in/foss4g2012 2
  • 4. Image Segmentation (Meanshift) Input image (CIELAB colour 5 Dimensional Input image(RGB) feature space space) sing Modes(local o lla p space c ect. Clustering ature ob j maximas) in fe form an i n ts o d e Po em to o n Objects formed Pruning (spatial Pruning (Spectral (backtracking the Bandwidth) Bandwidth) modes) Pruning ( Minimum region area) http://osgeo.in/foss4g2012 4
  • 5. Object attribution • Calculating textural, geometric and spectral features for the objects made in the Segmentation step in a feature vector. • Area, Perimeter, Roundness, Compactness, Centroid, Contrast, Coarseness, Direction, Roughness, Mean red, Mean green, Mean blue, std. deviation Red, std. deviation Green, std. deviation Blue. Segmented Image For every object in the image Initialize a vector having all 16 features Calculate value for every feature and save in the vector. http://osgeo.in/foss4g2012 5
  • 6. Classification • Mahalanobis Distance • Di,j2 = (x-µj)`S-1(x-µj) • The class which has the least Mahalanobis distance to the object i is the class of that object. Vectorization • Creates vector polygons for all connected regions of pixels in the object image sharing a common pixel value. • Polygon features are created on the output layer, with polygon geometries representing the polygons. • The class which has the least Mahalanobis distance to the object i is the class of that object. http://osgeo.in/foss4g2012 6
  • 7. The Input Orbview3 (4m) data of a part of delhi (500X500) The output of the objects with area less than 100. The shapefile(.shp) displaying the objects Output of object having area 100-200 and classified as building. http://osgeo.in/foss4g2012 7
  • 8. Performance Analysis Image Size Number of Running time objects (sec) 512 X 512 153 3.1 1024 X 1024 435 24.55 Table 1 : Image segmentation Image Size Number of objects Running time (sec) 256 X 256 49 0.249 512 X 512 193 1.133 1024 X1024 752 5.359 2048 X 2048 2965 31.60 4096 X 4096 11922 286.59 Table 2 : Object Attribution http://osgeo.in/foss4g2012 8
  • 9. Source Code • https://github.com/mohitkharb/Opticks_GSO C2012 • http://opticks.org/confluence/display/~mohit kharb/Workflow+of+the+pluggin • http://code.google.com/soc/ http://osgeo.in/foss4g2012 9
  • 10. Any Questions? http://osgeo.in/foss4g2012 10

Editor's Notes

  • #2: This paper describes a tool that implements Feature/Object Based This Image analysis for the Opticks remote sensing and image analysis software platform. These tools will partition remote sensing (RS) imagery into meaningful image-objects, and assess their characteristics through spatial, spectral and temporal scale. OSGEO-India: FOSS4G 2012- First National Conference "OPEN SOURCE GEOSPATIAL RESOURCES TO SPEARHEAD DEVELOPMENT AND GROWTH” 25-27TH OCTOBER 2012, @ IIIT HYDERABAD
  • #3: OSGEO-India: FOSS4G 2012- First National Conference "OPEN SOURCE GEOSPATIAL RESOURCES TO SPEARHEAD DEVELOPMENT AND GROWTH” 25-27TH OCTOBER 2012, @ IIIT HYDERABAD