Remote Sensing Data Processing in Sweden

• Main Author: Akerston, Ingvar S.
• Format: Text
• Language: unknown
• Published: Purdue University 1975
• Subjects: Sea ice
• Online Access: https://docs.lib.purdue.edu/lars_symp/97; https://docs.lib.purdue.edu/context/lars_symp/article/1098/viewcontent/1975_4B_15.pdf

Machine processing of remotely sensed data in Sweden is at present limited to one research group. Current activities encompass development of maximum likelihood classification procedures on a general purpose IBM 360/75 and later also on a special purpose PDP 11/40 with Ramtek GX 100 B Graphical display unit for interactive processing. Our Algorithm is adapted from the interesting modification of this method used by Robert H. Dye of the Bendix Corporation. In a sense this technique amounts to a "dynamic" determination of the optimum information yielding combination of a few wave length bands from a larger number of available spectral bands. Testing and evaluation of the analysis programs is done in an application to land vegetation mapping and the particular problem wed for these purposes is the identification of clear cut areas in forested regions. One of the output units in our special purpose image processing system is a novel Swedish color ink jet plotter, capable of generation hardcopy imagery in three colors about 10 - 16 gray levels each on plain paper or transparencies. It is a drum plotter with a registration capacity of 3 ⨯ 1024 ⨯ 1536 points in about 120 seconds on an area of about 20 CM ⨯ 28 CM. This project as well as other major remote sensing projects is sponsored in part by the remote sensing committee of the Swedish board of space activities. The research is mainly concentrated to three problem areas of particular interest VIZ land vegetation, sea ice, and oil pollution.