There are many difficult requirements imposed on block triangulation when it is applied on a practical basis. The main problem with aerial analytical triangulation is the error occurence and the necessity of reconstructing the block model. Because a certain number of operational errors always occur, the system of block triangulation is evaluated by the way it corrects such errors with certainly while at the same time reducing total working time. The authors of this paper have solved the problem by introducing the concept of block to block triangulation. With this concept, operational errors are kept in local and the volume of errors needing correction is minimized in total. Requirements for computerr use on practical basis are as follows: - 1. The error check method 2. The saving of computer costs (using only core memory-300 kc) To satisfy these requirements we have developed two concepts: 1. The concept of block to block triangulation 2. The reduction method on divided linear equation The system which the authors of this paper have developed has a cost/performance ratio level that is nearly comparable to the usual strip type solution method.
An approximate camera calibration method without using any photogoniometer or angle measurement was tried for the preliminary calibration of the taken camera. A flat terrain with vast quantity of canals, rivers and ditches was selected and the heights of water surface was measured with a precise plotting instrument at many points scattering over the stereo-model. After adjustments of model inclination and the effects of the Earth's curvature, the height discrepancies at each points from the datum plane was expressed as the function of bi-polar coordinnates, which has their poles at the principal points of the two photos. Assuming a polinomial form for the distortion, the residual bight discrepancies were computed and compared with the measured values And the distortion polinominals was corrected by iteration. Actual analysis was carried out graphically, with not so high accuracy which can not be expected from such a approximate method. Also, the determination of the principal points on the photo-plane was not tried because of the poor resultant accuracy which may be expected from such a approximate method.