Skip to main content

Visual Image Interpretation

Visual perception, i.e. looking at the environment, is an everyday experience for most of us. This ability has to be used for remote sensing as well. Image objects and structures have to be recognised and interpreted. Different factors influence the eye of the beholder, e.g. size, shape and position of objects as well as contrast and colour saturation of an image and the previous knowledge of the perceiving person.

Schema einer Sicheldüne

Crescentic dune (Barchan)

The shape is formed by constant wind with unchanged wind direction. The smaller ends of the sand dune are moving faster than the main dune and thus the characteristic semilunar structure is formed.

 

Satellitenbild einer Sicheldüne
Schema einer Längsdüne

Longitudinal dune (Seif)

Constant unidirectional wind forms the longish shape of the dune. The side facing the wind is wider and higher than the other side. These dunes can be several kilometres long.

 

Satellitenbild einer Längsdüne
Schema einer Sterndüne

Star dune

Changing wind directions lead to huge sand forms. The sand is reallocated again and again leading to a new formation.

 

Satellitenbild einer Sterndüne

Different shapes of dunes, schematic illustration vs. satellite image. The arrows stand for the main wind direction.


The figure above shows the different shapes of dunes. Knowing the ideal-typical scheme, one can find out that the satellite image shown above depict dune formations. Which kind of dune is prevailing can be found out at a second glance. This information can be used to derive the main wind directions.

 

How do we structure visual image interpretation?

Visual image interpretation is best-structured if organized in three stages:

  1. General areas (settlement / tree-covered / open space / water bodies)
  2. Similar areas  (fields / grassland / speciality crops / mixed)
  3. Single objects (plant species / fallow land / vegetation density)

Schema visuelle Bildinterpretation


Schematic illustration of visual interpretation by the example of Doha, capital city of Qatar (input data: NASA).

 

The procedure of image interpretation is carried out as seen above: First of all, one has to search for boundaries between different kinds of areas or colours. We have highlighted them in the second image. Then, finer structures, i.e. smaller image sections, are focused upon. The third image shows streets (yellow) or industrial spaces (violet) as examples.

 

You can try to visually interpret the images below yourself:

 

Besipiele visuelle Bildinterpretation

Left: Syr Darya, Kazakhstan, right: Minnesota, USA (© NASA)

 

How to obtain indirect information

The image below depicts the Cologne Cathedral (Kölner Dom) from a bird's eye view. At a first glance, we can collect information on shape, area and surroundings of the building. A second glance reveals the height of the object as well: The cast shadow of the object is used to calculate the height. It is apparent how the steeples and the nave vary in size; this would not be visible without the shadow.

 

Kölner Dom mit Schlagschatten


Aerial image of the Cologne Cathedral. The length of the shadow combined with the solar altitude can be used to determine the height of the cathedral indirectly. (© Google Earth).

 

In order to determine the height correctly, the angle of incidence of the sun has to be known, or the height of a reference object (like a fountain) as well as the length of its shadow. Then the height can be calculated, by using either a trigonometric function or the theorem of intersecting lines. Thus, indirect image interpretation supports quantitative and qualitative statements about the depicted objects.

 


Conclusion:

Visual image interpretation is a first analysis approach to remote sensing imagery. Here, the size, shape, and position of objects as well as the contrast and colour saturation are analysed. The height of objects can be determined by indirect visual analysis.