Abstract Users of the least squares algorithm (LS) coordinate transformation, usually focus on observation errors, but errors of the model coefficient matrix are ignored. Although total least squares algorithm (TLS) could take into account the errors of observations and the coefficient matrix, it cannot
solve the problem of gross errors effectively if we start with coordinate data and observations with gross error. Through the IGG I function, which gives the right of the observation to reduce the weights of the observation with gross errors, this article presents a kind of TLS algorithm based on reweighting iteration. During the three-dimensional coordinate conversion we can deal with the observations of gross errors effectively, and it is proven through example that the TLS algorithm based on reweighting iteration can get a more accurate three-dimensional coordinate conversion parameter
XU Bo,GAO Jingxiang,LI Zengke et al. The Application of the Total Least Squares Algorithm Based on Reweighting
Iteration to the Conversion of Three-Dimensional Coordinates[J]. jgg, 2015, 35(4): 693-696.
XU Bo,GAO Jingxiang,LI Zengke et al. The Application of the Total Least Squares Algorithm Based on Reweighting
Iteration to the Conversion of Three-Dimensional Coordinates[J]. jgg, 2015, 35(4): 693-696.