Ann M. Foster John Jones
The earliest known aerial photographs, from the mid to late 1920s, and resulted in the production of T-Sheets (Topographic Sheets) for the coasts and shorelines of south Florida. The T-Sheets are remarkably detailed, delineating features such as shorelines, ponds, and waterways, in addition to the position of the boundary between differing vegetation communities. If followed through time changes in the position of these ecotones could potentially be used to judge effects of changes in the landscape of the Everglades ecosystem, providing a standard by which restoration success can be ascertained.
The overall objective is to create a digital archive of historical aerial photographs of Everglades national park and surrounding area of the greater Everglades and south Florida. The archive will be in readily available Geographic Information System formats for ease of accessibility. Each set of photos will be broadly disseminated to client agencies, academic institutions and the general public via Open-File Reports and through the Internet.
Foster, Ann M. Briere, Peter R. Jones, John W. Van Arsdall, Carson
Neidrauer, C. J. Johnson, R. A. MacVicar, T. K. Perkins, W. A.
S. M. Davis and J. C. Ogden, editors
Gunderson, L. H. Park, W. A. Richardson, J. R. Mattson, J. E.
S. M. Davis and J.C. Ogden, editors
Havens, K. E. Carrick, H. J. VanZee, R
J. W. Porter and K. G. Porter, editors
Loxahatchee National Wildlife refuge U.S. Department of Agriculture, Natural Recources Conservation Service, Collier and Miami-Dade County offices South Florida Water Management District
Scanning: At each software initialization, settings will be established and / or verified. Scanning resolution may be determined by multiple factors, including the end needs of the user, processing constraints, storage space, and photo quality. Higher resolution photos (e.g. the 1987 photoset) may result in better output, but will require increased scan time, larger storage requirements (400-500 CDROMs) and increased processing and georeferencing time (see below). For example, we estimate that bringing the 1987 photoset to the rectification stage at a scan resolution of 300 dpi will take eight weeks.
A transparent template will be used to orient the photographs and eliminate the need to preview each individual image. Scanning will take approximately 3 minutes or longer per image. Following scanning the image will be archived. Files will be saved to an external storage medium. The original image file will be cropped to the area of interest as needed.
Rectification: Rectification will be accomplished via an ortho-rectification process using ERDAS OrthoBase or OrthoBase Pro software. This software depends on camera calibration reports that are readily available for later photosets, but not earlier ones. For the earlier photosets we will use a Non-Metric Camera Model. However, with knowledge of the camera (focal length, camera type) and the calibration reports, we will be able to extract the information necessary to construct a Frame Camera Geometric Model for the later photosets.
With a minimum number of Ground Control Points (GCPs) specified, the software automatically collects “tie” points. The tie points will be manually examined and those with obvious error eliminated. GCPs will be distributed across the area being rectified to minimize distortion. For the region being studied, good ground control points are rare and poor GCPs are better than no GCPs. Wherever possible we will use known GCPs and may collect field GPS readings for identifiable locations in earlier photosets that are extant today such as canals, levees, and structures. We will use Digital Ortho-photo Quadrangle Quarters (DOQQs) as the reference data set for rectifying each scanned image. An image(s) will be matched with the appropriate DOQQ. By the use of DOQQs, we can use good GCPs from one image to assist in rectifying an adjacent image with few or no GCPs. It is still critical to have a minimum number of good GCPs.
Sampling error will be assessed by use of the triangulation report produced by OrthoBase. This report provides the Root Mean Square Error (RMSE) for each group of images. We will not accept rectification results with a high RMSE. Rectification will be an iterative process as GCPs are added and/or removed until the RMSE is within the desired level.
The final phase of rectification involves resampling the image and exporting the resultant, corrected (rectified) image to GEOTIFF format. The GEOTIFF images will be archived to external media.
U.S. Department of the Interior, U.S. Geological Survey, Center for
Coastal Geology
Comments and suggestions? Contact: Heather
Henkel - Webmaster
Generated by mp version 2.7.19 on Thu Aug 14 10:11:00 2003