Over the last decade, the unprecedented population growth rates throughout much of the world and the subsequent need for information to support and manage that growth has put more pressure than ever before on geographical information system (GIS) and production mapping professionals. Businesses and planners need more up-to-date information than ever before, so much so that photogrammetry production “factories” can work 24/7 and still find it a challenge to keep up with the increasing volume of work.

The GIS professionals who typically rely on these data “factories” for assistance with data layer creation now find themselves standing in line. For busy GIS and mapping professionals, the solution to transforming imagery into reliable geospatial content more efficiently but without compromising accuracy lies in a process-driven workflow. This article discusses the process-driven workflow concept and the tools available to implement it effectively.

The Key to Increased Productivity

Subpixel positioning capabilities enable more accurate measurement of features.
Transforming imagery into usable data typically requires several processing steps. These steps are referred to as a workflow. It should be noted that following every step in this workflow is not necessarily required. Some scenarios might require certain steps but not others. However, typical scenarios will require every step in the workflow.

The first step in the process-driven workflow is to create a new project. This involves defining project properties such as the type of imagery that will be processed and the coordinate system used. This step also involves adding all of the raw data that will be processed to the project. When a project is created, the workflow begins.

A sensor model describes the properties and characteristics associated with the camera or sensor used to capture an image. Internal sensor model information describes the internal geometry, including focal length and lens distortion for aerial photographs. External sensor model information describes the position and orientation of each image as it existed when the imagery was collected. Without this information, value-added data layers such as oriented images, 3D feature datasets, Digital Terrain Models (DTMs), and orthorectified images cannot be derived from imagery.

Ground Control Points (GCPs) are used to establish a geometric relationship among the images in a project, the sensor model, and the ground so accurate data can be collected from the imagery. The GCP has three coordinates: x, y, and z, which are measured across multiple images. GCPs can be collected from existing vector files, orthorectified images, DTMs, and maps.

A tie point is a point with unknown ground coordinates but is visually recognizable in the overlap area between images. Tie points are used to position multiple images correctly, relative to one another. Automatic tie point collection uses digital image matching techniques to automatically identify and measure tie points across multiple images.

Block adjustment, which can include aerial triangulation, is essential to determining the information required to create orthophotos, DTMs, digital stereo models (oriented images), and 3D features. A block adjustment can obtain internal and external sensor model information, 3D coordinates of tie points, and additional parameters that characterize the sensor model. Most importantly, the results of a block adjustment can provide detailed statistical reports on the accuracy of data.

DTMs form the basis of many GIS applications and are vital for creating orthorectified images. To automatically generate a 3D terrain representation of the Earth and its associated geography, digital image matching techniques are used to automatically identify and measure the image positions of ground points appearing within the overlapping areas of two adjacent images. With this information, the accurate sensor model information from block adjustment is used to transform the image positions of the ground points into 3D coordinate information. After the automated DTM extraction process is completed, a series of evenly distributed, 3D mass points is located within the area and can be used to create a Triangular Irregular Network (TIN) or a raster DEM.