A Conjugate Gradient Approach to Least Squares Analysis of Cadastral Survey Data

Abstract

In theory the problem of maintaining the geometrical quality of a digital cadastral database can be solved if measurements are used instead of coordinates as the source data. This approach requires a very efficient least squares approach to solve the large amount of cadastral measurements. Although many least squares programs has been extensively developed, but they are designed specifically to optimize either geodetic or engineering problems. The nature of cadastral networks is unique compared that of the geodetic or engineering networks and hence require a new approach of the least squares computation. An iterative method known as conjugate gradient is being tested for cadastral applications.