Twin-camera fixed-base photography of tree plots: possibilities and accuracy of a system

Abstract

This paper investigates the stability and accuracy of a system using twin-camera fixed-base photography for the photogrammetric measurement of tree heights. Using basic matrix equations that relate photograph coordinates to ground coordinates, the general effects of errors of orientation parameters on tree heights are determined for the specific geometry of the system. It is shown that the most critical factor is the φ-inclination (the inclination of the camera along the axis perpendicular to the fixed base). The actual accuracy of the orientation parameters of each diapositive is then determined using a first-order analytical plotter (BC-1), with five stereopairs of tree plots taken at a nominal scale of 1:800. Because of the small base–height ratio, the standard deviations of the orientation parameters (ω, φ, κ) along the X, Y, and Z axes of each camera are high: between 10cc and 20cc (1cc = 0.01 grade). However, the cumulated effect of all the factors gives an accuracy within 30 cm for a tree height of 10 m. Compared with the effect of the relative orientation error in φ, the other factors have an almost negligible effect. Thus, important uncertainties concerning a practical system of deriving quantitative information from tree plots by photogrammetric means are clarified. Directions of further development are also indicated.