Investigation into Self-calibration Methods for the Vexcel UltraCam D Digital Aerial Camera

This paper provides an investigation into the camera calibration of a Vexcel UltraCam D digital aerial camera which was undertaken as part of the EuroSDR Digital Camera Calibration project. This paper will present results from two flights flown over a test site at Fredrikstad-Norway using established camera calibration techniques. Furthermore, it proposes an alternative approach. The "new" multi cone digital camera systems are geometrically complex. The image used for photogrammetric analysis is made up of a number of images produced by a cluster of camera cones and possibly various groups of CCD arrays. This produces a resultant image which is not just based on traditional single lens/focal plane camera geometries, but depends on the joining of images from multiple lens (different perspectives), handling groups of focal planes and the matching of overlapping image areas. Some of the requirements from camera calibration such as stability can only be determined through long-term experience/research and some can be determined through investigation and short-term research such as the calibration parameters. The methodology used in this research for assessing the camera calibration is based on self-calibration using the Collinearity Equations. The analysis was undertaken in order to try to identify any systematic patterns in the resulting image residuals. By identifying and quantifying the systematic residuals, a new calibration method is proposed that re-computes the bundle adjustment based on the analysis of the systematic residual patterns. Only very small systematic patterns could be visually identified in small areas of the images. The existing self-calibration methods and the new approach have made a small improvement on the results. The new calibration approach for the low flight has been particularly beneficial in improving the RMSE in Z and reducing image residuals. However, the method was less successful at improving the high flown results. This approach has shown that it has potential but needs further investigation to fully assess its capabilities.