3 ECTS credits
75 u studietijd
Aanbieding 1 met studiegidsnummer 4015683ENR voor alle studenten in het 1e semester met een verdiepend master niveau.
The course gives an overview of standard methods in (semi-) automated interpretation of earth observation data for land-use/land-cover mapping. The methods discussed in the theoretical part of the course are applied in practical case studies to be carried out by the students using remote sensing software. Results of the practical work are summarized and discussed in a series of scientific reports to be produced individually by each student.
1. Earth observation systems
a. Basic concepts
b. Sensor characteristics: low-, medium- and high-resolution multispectral sensors, hyperspectral sensors
c. Selecting an appropriate sensor
2. High-resolution models for image interpretation
a. High-resolution versus low-resolution scene models
b. Distribution of thematic classes in feature space
c. Parametric classification methods: multi-dimensional class distribution parameters, the multi-variate Gaussian distribution, the minimum-distance-to-means classifier, the maximum-likelihood classifier
d. Non-parametric classification methods: nearest neighbour classifiers, the multiple layer perceptron, decision tree classifiers
e. Image classification accuracy: overall measures of accuracy, class-specific measures of accuracy, Kappa analysis
3. Change detection methods
a. Introduction: impact of sensor characteristics and environmental conditions
b. Visual change detection: false color composite image generation
c. Multidate composite image change detection: multidate image classification, principal component analysis
d. Image differencing
e. Post-classification change detection methods: post-classification comparison change detection, use of a binary change mask, cross-correlation change detection
f. Spectral vector change analysis
Required prior knowledge: basic knowledge multivariate statistics
After successful completion of this course the student should:
- have a good knowledge of the characteristics and the use of important sensors for terrestrial earth observation;
- be familiar with a variety of techniques for image classification, their potential for land-use/land-cover mapping, and the conditions that should be fulfilled for applying these techniques;
- master alternative methods for change detection and be able to choose a proper method depending on the type of application;
- be able to apply standard earth observation techniques for producing land-use/land-cover information at different scale levels;
- be able to assess the accuracy of a classifier and compare classification results obtained with alternative methods for image interpretation.
De beoordeling bestaat uit volgende opdrachtcategorieën:
Examen Schriftelijk bepaalt 50% van het eindcijfer
WPO Verslag bepaalt 50% van het eindcijfer
Binnen de categorie Examen Schriftelijk dient men volgende opdrachten af te werken:
Binnen de categorie WPO Verslag dient men volgende opdrachten af te werken:
Students cannot pass for this course if not all scientific reports have been submitted. Students have to obtain a score of minimum 8/20 on both the theoretical and the practical part of the exam to pass for this course. When the score is below 8/20 for either the theory part or the practical part, the total score for this course will be set equal to the lowest score of both partial evaluations.
Deze aanbieding maakt deel uit van de volgende studieplannen:
Master in de geografie: Standaard traject
Master of Geography: Standard track (enkel aangeboden in het Engels)
Educatieve master in de wetenschappen en technologie: biologie (120 ECTS, Etterbeek)
Educatieve master in de wetenschappen en technologie: geografie (120 ECTS, Etterbeek)
Educatieve master in de wetenschappen en technologie: chemie (120 ECTS, Etterbeek)
Educatieve master in de wetenschappen en technologie: wiskunde (120 ECTS, Etterbeek)
Educatieve master in de wetenschappen en technologie: ingenieurswetenschappen (120 ECTS, Etterbeek)