| dc.description.abstract | The principal point has long been regarded as one of the fundamental parameters in camera calibration. In the age of film based aerial and large format terrestrial cameras, the principal point could be located by a variety of techniques with a certainty of ±10 mm (Carman and Brown, 1961) and this was considered sufficient. However, aerial cameras were precision, purpose built, expensive pieces of equipment where the assembly was painstaking and the location of the principal point measured to a known tolerance. In the digital era, photogrammetrists, and many others, are using cameras which have not been specifically designed or built for photogrammetry. For these cameras there is no requirement for the manufacturers to position the lens in a pre-defined location relative to the image sensing plane or for the lens manufacturer to align the lens elements precisely. In fact, deviations from the centre of the sensor can be a considerable percentage of the extent of the sensor (up to 10 per cent for some zoom lenses (Burner, 1995)). This paper discusses the development of methods of obtaining the location of the principal point, considers the relationship between the principal point and other parameters in the functional model, and shows how the location of this point can be estimated with and without recourse to autocollimation methods | en |
