Map-to-Map Coordinate Conversions

Producer Field Guide

Producer Field Guide

There are many instances when you may need to change a map that is already registered to a planar projection to another projection. Some examples of when this is required are as follows (Environmental Systems Research Institute, 1992):

  • When combining two maps with different projection characteristics.
  • When the projection used for the files in the data base does not produce the desired properties of a map.
  • When it is necessary to combine data from more than one zone of a projection, such as UTM or State Plane.

A change in the projection is a geometric change—distances, areas, and scale are represented differently. Therefore, the conversion process requires that pixels be resampled.

Resampling causes some of the spectral integrity of the data to be lost (see the disadvantages of the resampling methods explained previously). So, it is not usually wise to resample data that have already been resampled if the accuracy of data file values is important to the application. If the original unrectified data are available, it is usually wiser to rectify that data to a second map projection system than to lose a generation by converting rectified data and resampling it a second time.

Conversion Process

To convert the map coordinate system of any georeferenced image, ERDAS IMAGINE provides a shortcut to the rectification process. In this procedure, GCPs are generated automatically along the intersections of a grid that you specify. The program calculates the reference coordinates for the GCPs with the appropriate conversion formula and a transformation that can be used in the regular rectification process.

Vector Data

Converting the map coordinates of vector data is much easier than converting raster data. Since vector data are stored by the coordinates of nodes, each coordinate is simply converted using the appropriate conversion formula. There are no coordinates between nodes to extrapolate.