PROPAGATION OF VARIANCE AND COVARIANCE

In a direct least-squares solution, the variance-covariance matrix of the unknown parameters may be computed by multiplying the inverse of the coefficient matrix of the normal equation by the variance of unit weight. In an iterative least-squares solution, which is generally applied in problems of analytical photogrammetry, this formulation is theoretically valid only if the corrections to all the approximations become zero in the last iteration. Experimental evidence showed that this formulation could not detect any rapid accumulation of systematic effects caused by random errors in the measured parameters. Experimental results also showed that this formulation could provide reliable estimates on the RMS errors of the computed parameters if the correction parameters converge to a value which is less than the computed RMS errors. /Author/

  • Availability:
  • Supplemental Notes:
    • Presented at the Annual Convention of the American Society of Photogrammetry, St. Louis, March 1974.
  • Corporate Authors:

    American Society of Photogrammetry

    105 North Virginia Avenue
    Falls Church, VA  USA  22046
  • Authors:
    • WONG, K W
  • Publication Date: 1975-1

Media Info

Subject/Index Terms

Filing Info

  • Accession Number: 00095961
  • Record Type: Publication
  • Files: TRIS
  • Created Date: Jul 2 1975 12:00AM