GMFM350 Big data and machine learning in remote sensing

Credits (ECTS):10

Course responsible:Misganu Debella-Gilo

Campus / Online:Taught campus Ås

Teaching language:Engelsk, norsk

Course frequency:Every year

Nominal workload:Total structured teaching: 100 hours. Exercise work without direct supervision: 96 hours. Individual study: 50 hours.

Teaching and exam period:Autumn parallel

About this course

Lectures: Characteristics and processing chains of remote sensing data. Big data concepts and tools in remote sensing. Machine learning (ML) and deep learning (DL) methods and algorithms used in remote sensing. ML and DL in remote sensing image classification, segmentation and processing (correction and enhancement). ML and DL methods for analyzing high dimensional remote sensing data such as laser scanning data, image time series, and hyperspectral images.

Exercises:1: Big data in remote sensing: image collections, data cube, time series and their computation. 2: Multi-spectral image classification and segmentation using Machine learning methods. 3: Satellite image time series, lidar data and hyperspectral image classification using machine learning.

Learning outcome

After completing the course, the students will have obtained substantial insight, technical understanding and practical experience with machine learning and deep learning algorithms used in the processing and analysis of various types of remote sensing images particularly in image preprocessing, classification, segmentation, and change detection. Additionally, the students will have been acquainted with the understanding and basic competence of handling remote sensing big data. Python is the workhorse of this course and basic competence in Python is requires.
  • Learning activities
    The lectures introduce and refresh theoretical backgrounds relevant to the use of machine learning algorithms in remote sensing by characterizing remote sensing images, their processing chains and exploring applicable machine learning and deep learning algorithms. The exercises provide hands-on experience with the application of machine learning in different uses of remote sensing data.
  • Teaching support
    Teaching support will be given primarily in connection with that part of the structured teaching that is set aside for exercise guidance. It will also be possible to communicate directly with the course teacher by appointment during office hours.
  • Prerequisites
    GMBB100/GMFM100 or similar (e.g. MINA305), and DAT200 or similar
  • Recommended prerequisites
    GMBB201/GMFM200 or GMBB300/GMFM300, INF200
  • Assessment method
    Portfolio assessment based on project assignments (accounts for 50% of the grade) and written exam (counts for 50% of the grade). Grading scale A-F.

  • Examiner scheme
    The external examiner collaborates with the internal examiner in designing examination tasks and guidelines. The external examiner verifies the internal examiner's assessment of a random selection of candidates periodically, as a part of the calibration process according to the institute's grading guidelines.
  • Mandatory activity
    Obligatory Exercises
  • Teaching hours
    Lectures: 30 hours. Exercises: 60 hours. Excursion: 8 hours. Continuous assessment: 2 hours.
  • Reduction of credits
    None
  • Admission requirements
    Special requirements in science