This technique can be used to provide higher resolution and more detailed GIS-type information as the user zooms in, but using a limited number of images without having to always be retrieving files from a server.
Typical implementations provide one or two high resolution "basemap" images which contain "background", and the same number of one of the overlays, usually containing roads and/or geo-political boundaries.
The configuration file looks like this:
The file_of_filenames is simple....
The high_res_zoom parameter lists the resolution multipliers for each image. In this case, "2" means that the image CHI-2.png is 2x the size of the original image. The "4" means that the 2nd named image (CHI-3.png) is 4x the size and resolution of the original image. In all cases, the projections and geographic coverage must be identical for all images.
We suggest you zoom in on a spot along the lake shore, southeast of "Chicago"
-- on the low-resolution image, it looks like a small
island closer to Gary. (This is, Indiana Harbor...).