LAND-USE MAPPING BY MACHINE PROCESSING OF LANDSAT-1 DATA

Data on computer-compatible tapes of LANDSAT-1 MSS Frame I-1483-15132, of August 30, 1973, were analyzed to generate a land-use map of a portion of the City of Virginia Beach, Virginia. Bands 4,5,6, and 7 were used in the supervised approach with the LARSYS software system of Purdue University. A Land-Use/Land-Cover map at a scale of 1:24,000 was obtained. Major functional classes delineated were "Urban," "Agricultural," "Wooded," "Water," "Wetland," and "Bare Land." Twenty-four subdivisions of these classes were spectrally separable. Some of these were subdivisions of Level II of the Inter-Agency Steering Committee on Land-Use Information and Classification. Extent of man-made structures, unit density, and degree of weathering were important in subdividing the Urban class. In the Agricultural class, crop type was the most important factor. Vegetation types, age, and thickness of undergrowth were important in Woodland and Wetland areas; the depth, turbidity, and degree of algal and other pollution determined Water subclasses. Bare Land consisted mostly of beaches and mudflats. Performances of 97.9 percent for training field, 93.9 percent for test field, 97.0 percent for training class, and 92.1 percent for test class were obtained. Construction activities caused some misclassifications. The LARSYS software system was found to be useable in areas with complex land uses. The production of highly accurate functional land-use maps with the machine processing approach was found to be feasible.

  • Availability:
  • Corporate Authors:

    American Society of Photogrammetry

    105 North Virginia Avenue
    Falls Church, VA  USA  22046
  • Authors:
    • Odenyo, VAO
    • Pettry, D E
  • Publication Date: 1977-4

Media Info

Subject/Index Terms

Filing Info

  • Accession Number: 00164068
  • Record Type: Publication
  • Files: TRIS
  • Created Date: Nov 9 1977 12:00AM