Section: Classification and Feature Extraction | Remote Sensing and Image Analysis | Z_GIS

Main course page

  • General

    • Introduction Forum
    • Development feedback Forum

      Please post all suggestions and corrections for this module. Refer to specific sections whenever your posting applies to content of a specific section.

    • Learning objectives (module level):

      • appreciate the value of remote sensing imagery as a core element for a(ny) Digital Earth
      • explain and communicate the fundamentals of remote sensing including physical principles and characteristics of platforms and sensors
      • gain experience with at least two different cloud platforms for remore sensing image analysis - e.g. Sentinel Hub and ArcGIS Image for ArcGIS Online
      • modify multi-band image display according to information needs and application domains
      • visually integrate imagery with geospatial displays, products and apps
      • access, manage, explore and integrate remote sensing imagery into geospatial workflows and products
      • understand and adequately apply simple classification methods, starting from indices
      • work with raster analysis methods, including pre- and post-processing 
      • implement simple change detection and monitoring tools 
      • be aware of advanced topics of remote sensing data management and analysis in general terms 

Imagery to thematic layers
Monitoring and detecting change

Classification and Feature Extraction

  • Classification and Feature Extraction

    • Learning objectives: 

      • know the basics of image classification, understand and apply multivariate image classification models, i.e. unsupervised vs. supervised classification
      • assess the quality and accuracy of image classification results
      • visualize and share thematic image processing outputs
      • understand and apply (pre-configured) basic machine learning approaches
      • interpret results from object-based image analysis (OBIA) feature extraction

    • Section overview 

      Pop out | Video

    • Image Classification using ArcGIS URL

      Learning path based on ArcGIS Pro - recommended if this platform is readily available. 

    • Esri Academy: Introduction to Image Classification (Web Course) URL

      Supervised pixel classification - Unsupervised pixel classification - OBIA 

    • Classification: Decision Trees URL

      EO4GEO Lecture on this popular non-parametric classifier (advanced topic). >Presentation 

    • Classification: Random Forests URL

      EO4GEO Lecture introducing a machine learning classifier with a capability to obtain good classification results with a relatively low number of training samples and reliance on a reduced number of user-defined parameters (advanced topic). >Presentation

    • Object-based Image Analysis – An Introduction URL

      EO4GEO Lecture providíng an introduction to concepts and methods of Object-based image analysis. >Presentation

    • Artificial Intelligence for Earth Observation URL

      A webinar addressing the use of AI algorithms in EO applications. Introducing AI concepts, from machine to deep learning, from unsupervised to supervised methods. Furthermore, applications in object detection, semantic segmentation, classification, clustering, and data augmentation will be showed.

    • Activities: Page

      Tasks to complete this section

    • Assessment: Page

      Check out these questions and exercises

Imagery to thematic layers
Monitoring and detecting change