Skip to main content icon/video/no-internet

Aerial Imagery: Interpretation

Aerial images are photographic or digital images captured from a position above the ground. These images provide visual information over a wide area that is difficult to obtain at the ground level. Interpretation is the act of examining these images for the purpose of identifying, characterizing, and measuring objects and phenomena and judging their significance. Interpretation is not considered to be an exact science, and the interpretations are often based on the probability that an object is identified accurately. Successful interpretations depend on practical training and experience, disciplined approaches using knowledge of aerial imagery and application of that knowledge, as well as the inherent talents of the interpreter. Computer technology can aid in interpretation, but final identification by a knowledgeable human is often necessary. Interpretation is key for linking remotely sensed information to ground information on Earth's surface. This entry discusses the fundamental approaches to aerial imagery interpretation.

Interpretation Tasks

There are four tasks in image interpretation, listed in order of increasing sophistication: detection, identification, measurement, and problem solving. Detection is the lowest order and determines the presence or absence of an object or phenomenon on the ground. An example is the detection of buildings or roads in an urban environment. Identification is more advanced and involves the labeling of detected objects, such as schools or highways. Measurement allows the quantification of ground objects measured directly from the imagery, such as the number and area of buildings or the length of a highway. Problem solving is the most complex task and uses information from the first three tasks for higher-level identification, such as commercial building density.

Types of Aerial Imagery

The type of aerial imagery used can determine the level of interpretation. Aerial imagery is captured at varying spectral wavelengths, from panchromatic imagery across visible light wavelengths, to color infrared imagery, which captures energy in the visible and near-infrared wavelengths (see Figure 1), to hyperspectral imagery, which divides captured energy into many narrow bands of information. Figure 1 shows a typical color infrared image, where green wavelengths are represented by blue in the RGB spectrum, red wavelengths are green, and near-infrared wavelengths are red. The wavelengths at which imagery is captured and viewed can affect successful interpretation. For instance, in Figure 1, healthy vegetation can be easily identified as red, because high amounts of energy are reflected by plants in the near-infrared wavelengths, shown here as the red band. Imagery is also captured at a number of angles, from directly above to oblique. Stereoscopic imagery, with multiple oblique views of the same landscape, can aid interpretation by giving the image a three-dimensional perspective. The spatial resolution of imagery can also vary based on the type of camera used, as well as the height of the camera from the ground. The spatial resolution refers to the smallest possible feature that can be distinctly detected from its surroundings. Imagery captured at a higher altitude will have coarser spatial resolution than the same imagery from a lower altitude.

Interpretive Elements

There are several elements that can be interpreted from aerial imagery. The most basic elements can be easily interpreted using a computer. However, computers have difficulty with more complex elements, which require expert knowledge from a human for interpretation, where cognition can vary from person to person.

...

  • Loading...
locked icon

Sign in to access this content

Get a 30 day FREE TRIAL

  • Watch videos from a variety of sources bringing classroom topics to life
  • Read modern, diverse business cases
  • Explore hundreds of books and reference titles

Sage Recommends

We found other relevant content for you on other Sage platforms.

Loading