High Spatial Resolution Digital Imagery
89
9
CONCLUSIONS AND RECOMMENDATIONS
Each of the four study sites demonstrated the potential use of high spatial
resolution digital remotely sensed imagery for a specific application. Successes
and shortcomings were presented for each application. The following broad rec-
ommendations provide guidance in the effective and efficient application of high-
resolution imagery.
The total size of the project site or sites, measured as the total study area
(acres or hectares), will be one of the primary factors in determining the
type of high-resolution imagery to acquire. For areas less 50 km2, air-
borne imagery may provide the most cost effective image source. For ar-
eas of interest approaching 100 km2, high-resolution satellite data will be
the cheapest solution.
When contracting the acquisition of airborne digital imagery, either
multispectral or hyperspectral, ensure that the contractor provides accu-
rate and comprehensive metadata describing the precise acquisition pa-
rameters, as well as the tools and techniques used to radiometrically and
geometrically post-process the images. Generally, expect higher image
post-processing costs for hyperspectral imagery versus multispectral im-
agery. If the contractor is to develop a map product, require an exhaus-
tive report that documents all ground truth data collection procedures and
image processing routines, including a statistically valid accuracy as-
sessment.
Ensure that an adequate volume of reliable ground truth data is collected
and analyzed to meet the objectives of the mapping component of the
overall project. Roughly plan on equal dollar amounts for image acquisi-
tion and ground truth data collections costs. Image post-processing, im-
age classification, accuracy assessment, and report generation costs will
be determined by the overall size of the project site.
9.1
Based on the overall quality of the imagery in terms of spatial resolution, ra-
diometric balance, geometric accuracy, and land cover classification, the objec-
tives of this pilot project were met. The ArcView shape files should provide the
Poplar Island wetlands design team useful information concerning the areal cov-
erage, distribution, and configuration of the three primary surface features (i.e.,
water, intertidal zone, and vegetated wetlands). Of the 10 reference wetland areas
imaged, eight sites provided land cover class maps.