Review of Digital Image Orthorectification Techniques

Dr. F. I. Okeke
Department of Geoinformatics and Surveying,
University of Nigeria, Enugu Campus, Nigeria


Introduction
Aerial photos and satellite images do not show features in their correct locations due to displacements caused by the tilt of the sensor and terrain relief. Orthorectification transforms the central projection of the photograph into an orthogonal view of the ground, thereby removing the distorting effects of tilt and terrain relief. Orthorectification is the process of transforming raw imagery to an accurate orthogonal projection, as against the perspective projection of the raw image (Fig 1). The product of orthorectification process is orthoimage or orthophotos. Without orthorectification, scale is not constant in the image and accurate measurements of distance and direction cannot be made.


Fig.1: Orthogonal vs. perspective projections


In order to orthorectify imagery, a transformation model is required which takes into account the various sources of image distortion generated at the time of image acquisition. These include:

  • Sensor orientation
  • Topographic relief
  • Earth shape and rotation
  • Sensor orbit and attitude variations
  • Systematic error associated with the sensor
The required geometric parameters regarding sensor orientation at the time of image acquisition are determined through information on the sensor model, Ground Control Points (GCPs), and platform orbital or flight data (position, velocity, orientation).

Because orthoimages are planimetrically correct, it can be an effective tool for use in resource management, municipal planning, cadastral mapping, and geographic information systems (GIS). It is not only true in scale and area, but like a conventional aerial photograph it is easily interpreted. A forester in the field of resource management can designate and outline forest-type boundaries on an orthophoto directly.

However, the conventional orthorectification processes do not take into account objects that mathematically can be described like buildings, bridges, trees etc. Such objects remain in perspective views in the resulting orthophotos and are distorted from their true positions. Distortions show as, for instance, leaning buildings and bent bridges. Some interesting information from ground features like streets and other objects are hidden from the user of the orthoimage. Furthermore the superimposition of vector data is nearly impossible. If Digital Surface Models (DSMs) describing the mentioned objects that caused the displacement are used for the orthorectification processes, the displacements can be corrected and the results are called “True Orthoimages”.

This article however reviews only some of the conventional digital orthorectification techniques.

Orthorectification algorithms
Generally there are two classes of rectification approaches. The parametric and the non-parametric approaches (Hemmleb and Wiedemann, 1997). Whereas for the parametric approach the knowledge of the interior and exterior orientation parameters is required, non-parametric approaches require just control-points. Non-parametric approaches include polynomial transformation, and projective transformation. A comprehensive comparative study of orthorectification approaches can be found in Novak (1992).


Page 1 of 3
Next