QuickDraw Data Structures for Image Processing
Abstract
Standard binary data formats are currently used to import and export satellite images to geographic information-and image-processing systems. These data structures provide a standard sequential method to read and write large volumes of information in a semicompressed format. While the binary structure is adequate for strict import of image data, it is poorly adapted to fast image- processing at the microcomputer level. In this study, new data structures are investigated that use operating codes to quickly convert raster binary image data and vector overlay files into a high-speed graphical language for efficient display and processing. Binary data is converted into 'picture handles' of variable size and resolution using 2-byte operating codes to symbolize the graphical process. As a result, images are drawn as objects that may be coupled as independent vector components in multiple bit planes. The bit planes may be specified for each pixel to support 24- and 32-bit color of both raster and vector data. The efficiency of these structures allows the user to display 1024 x 1024 scenes in multiple overlapping windows using simple Cut/Copy/Paste commands. In addition, the use of operating codes allows an analyst to quickly store and retrieve archived images in their compressed form using simple 'scrap manager' techniques. Early results for this technique indicate that converted vector overlays may be compressed by a factor of 8 and SPOT images (depending on scene diversity) by a factor of 2. More significantly, images that typically require 4 min to load from binary may be displayed in fractions of a second using the new display method and resultant operating codes. (RH)
Document Details
- Document Type
- Technical Report
- Publication Date
- Apr 01, 1989
- Accession Number
- ADA208945
Entities
People
- Harlan L. Mckim
- Perry J. Lapotin
Organizations
- Cold Regions Research and Engineering Laboratory