Total least squares with application in geospatial data processing

The measurements in geospatial surveys always contain random errors. In this paper, the Total Least Squares (TLS) method, which accounts for the errors both in the observation vector and design matrix, is applied in the triangulation geospatial surveying adjustment. Two frequently used TLS algorithms, which are the singular value decomposition (SVD) based TLS algorithm and the iterative TLS algorithm are utilized. Comparisons and analyses of the TLS algorithms with LS algorithm are made in terms of observation residuals, estimated variance component and estimated errors of the adjusted parameters in the adjustment model.