Automated cartographic feature attribution using panchromatic and hyperspectral imagery : DARPA/APGD yearly report 1998-1999

Abstract : The first part of this report describes experiments in the block adjustment of linear pushbroom sensor imagery incorporating object-space straight-line constraints. Both polynomial and interpolative platform models were tested; the polynomial model generally performed better than the interpolative model without lines, but not as well as the interpolative model with straight line constraints. The greater flexibility of the interpolative model makes it better able to describe complex platform motion and to use the geometric strength given by the straight-line constraints, at the expense of increased sensitivity to uneven point distributions or bad points. The second part described current quantitative classification results for man-made and natural materials using 14 surface material classes over selected test areas with Fort Hood. Radiometric effects due to changing solar illumination and atmospheric conditions during the acquisition issues also are discussed.