Automatic Cartographic Feature Identification.

Abstract : The objective of the study efforts presented in this report was to identify and test automatic identification techniques whereby descriptor data is entered into the digital record for selected features and additional descriptor data for given sets of features is derived by computational techniques. This study was divided into three phases. In the first phase, general background and knowledge of the problems was gained, and potential solutions were identified. This was accomplished by site visits to DMA facilities, literature review, and discussions with experts in cartography and related areas. Section 2 of the report discusses this phase in greater detail. The second phase was concerned with experimental testing of techniques identified as having the greatest potential payoff. Section 3 of the report details the experiments conducted, their significance and recommended further testing in these areas. Section 4 contains more detail on suggested further research based upon the experimental results. The third and final phase was concerned with the development of a conceptual integrated system for the digital processing of cartographic data. Section 5 presents Andrulis' general concepts for the design of such a system. (Author)