Material
 

Topic 1: Electromagnetic Radiation

Theory and Learning

In this subtopic we will study the what happens when the EMR radiation Interacts with a translucid object as the atmosphere. This interaction occurs both in the path down from the Top of the Atmosphere (TOA) to the ground and in the path up (after reflection or thermal emission of the Earth) from the ground to the TOA where the satellite sensor is located.

• For solar radiation in the SW and LW the path is double (TOA to the Ground + reflection + Ground to TOA where the sensor is)
• For earth radiation in the LW, the path is single (Ground to TOA where the sensor is)

The atmosphere affects the radiation in both paths. The understanding of this interaction is the basis for atmospheric correction algorithms. Here we only explain this basis.

Topic 2: Sensors and Image Characteristics

 A "system" is a composition of "components" that organized in a "flow" are utilized to obtain products usable for different purpose.

In this unit we look at the "Remote Sensing System" as a flow of activities to produce images applicable in many areas of research. We stress more in the image characteristics and properties than in the engineering of the system, however a view of all those components will be presented.

Learning Objectives

• Differentiation: Sensors on board on platforms.
• Classification of sensors: Active and passive sensors
• Image constraints: Classification of imagery capability:
  • Spatial resolution: ground sample resolution, ground spatial resolution. Scale is meaningless in RS.
  • Radiometric resolution.
  • Revisiting time or temporal resolution.
  • Overpass time.
  • Expected lifetime.
• Criteria for image selection: cloudiness and angular distortions
• Summary of available missions, platforms and imagery (selection to specializations)

Topic 3: Sensor Calibration

Landsat 8 (NASA) carries sensors for Earth Resources research of last generation. The sensors operate in moderate to high resolution, in many bands and the data acquisition is free of charge. In this lecture the steps for image acquisition and image calibration (going from Digital numbers to Radiance and reflectance or brightness temperature in case of thermal images will be described.

Learning Objectives

At the end of this lecture the student should be able to:

• Understand the sequence of calibration for a generic sensor in the shortwave.
• Understand the sequence of calibration for a generic sensor in the longwave.
• Be able to explore the metadata file in order to obtain calibration coefficients, geometric information and datum needed to calibrate the images
• Apply the process to Landsat 8 OLI and TIRS.
• Use the raster calculation interface in QGIS

Topic 4: Coordinate Systems

Learning Objectives

At the end of this unit, the students should be able to:

• Understand the basics on Coordinate Systems
  • Ellipsoids and Horizontal Datum: many to one and one to many.
    • Country coordinate systems
  • Projections
  • Coordinate conversions

Topic 5: Georeferencing and Geocoding

Learning Objectives

At the end of this unit the student should be able to:

• Understand image distortions and displacements.
• Regarding Georeferencing: (2D approaches)
  • Perform a manual georeference over an image:
    • Understand planar interpolations between coordinates in X, Y and (X,Y)
    • Give coordinates from a map or GPS to an image
    • select adequate transformations as related to a certain type of imagery: conformal, affine and polynomial.
• To correct a distorted imagery to get it in the right projection and north oriented: From georeferencing to geocoding.
  • Matching images with maps.
  • Process of resampling.

Topic 6: Color, Enhancement and filtering

The images produced in a single band of a sensor needs to be "accommodated" to the human perception in order to be interpreted correctly. A rather "obscure image" of a low reflectance area needs enhancement to be investigated. Images that are taken away from the visible need to be displayed on the screen that works only with blue, green and red leds. This process creates images that are not natural for humans and we need to learn how to interpret them.

This topic explains the methods to adequately display and interpret sensor bands of Earth for specific applications.

Learning Objectives

At the end of this unit, the student should be able to:

• Understand the additive and substractive theory of colors.
• Apply the theory of Colors to image bands of satellites in the Shortwave and understand image composition
• Select 3 bands out of all the bands produced by a sensor, display a False Color Composite and understand the colors to a level that allows understanding and interpretation of the image.
• Explain and apply 2 major options to enhance image
  • histogram operations (linear and histogram equalisation) and
  • filter operations consisting on:
    • noise reduction,
    • edge detection and
    • edge enhancement) for visualization and digital image classification. (level 1, 3)
• Understand the two main types of image filters and its use.

Topic 7: Basis of DIC, data preparation , training samples and clustering

The Earth surface is rather heterogeneous. What is covering the ground is called "Landcover". Landcover maps are built after a classification of a remote sensed image, so the legend of such a map could be something like "Water", "Grass", 'Bare Soil", "Forest", and many more.

Despite the heterogeneity, there are patches of land that are covered by the same type of unit. These patches have similar reflective properties. So for instance, if we know which are the reflective properties of a "Grass" we could find in the image all grasses, with some image operation. To do that we need to train the Remote Sensing software to "recognize" the specific set of reflective properties of the "Grass". That is done by selecting pixels containing "Grass" and letting the system retrieve the reflectance of "Grass" in all bands that later will be used to detect all grasses in the image. This training of the computer is the first step in the DIC. 

The user repeats the process for other land covers, till the needs of the project is satisfied.

Topic 8: Image Classification Techniques, assessment and control

The classified image has to be verified. Normally a great number of pixels in the image are either not classified or classified wrongly. Some pixels may have similar spectral response to two or more units and then the statistical classification algorithm might fail in classified the pixel correctly.

In this "classification assessment" lecture two procedures of assessment of classification will be discussed:

1. The "Confusion Matrix" where the pixel samples taken by the user during the training part will be verified to belong to the right unit after the classification of the image and
2. The cross check of the classification with real information obtained after fieldwork.

Topic 9: Applications of Remote Sensing in Water related topics

This is a presentation on (some) current topics of research at ITC WRS department and parter departments. It gives you a glance on what we are doing and your potential field of research for the MSc degree.