Integration of cadastral survey plans into a computerized land information system
MetadataShow full item record
AbstractA primary source of data for Land Information Systems (LISs) is the cadastral survey plan. An automated system for entering, analyzing, and integrating the information shown on these plans is developed in this thesis. Existing systems are evaluated on the basis of their speed, robustness, completeness and simplicity, and are found to have serious limitations. The design of this new system begins with an enumeration of the types of entities that appear on cadastral plans and their relationships with each other. Rules are developed for the automatic detection and modelling of these relationships. This includes the formation of the polygons defining the lots, blocks, etc. shown on the plan, and the detection of constraint conditions, such as points that must lie on a straight line. Constrained least squares is presented as a powerful tool that uses not only the observations in the calculations, but also the topological relationships, for the calculation of the positions of the points on the plan. Data errors are detected using a combination of topological rules and statistical analysis. This ensures that the data entered into the LIS are virtually error free. The proposed system is implemented in a prototype program and the results of testing using real cadastral survey plans are presented. The implications of using this program in an existing production environment are examined and it is expected that significant productivity improvements and cost savings can be realized.
Bibliography: p. 120-125.