Preliminary tests of camera self-calibration by wavelet-based additional parameters

Camera calibration is an important issue in photogrammetry. Self-calibration by additional parameters is the most famous method to calibrate various types of cameras. Traditional algebraic polynomial additional parameters were proposed for analogue single-head camera calibration, and they are later also applied to diverse digital cameras. However, many different types of metric and non-metric cameras are widely used in photogrammetry and computer vision. Traditional algebraic polynomial additional parameters might not be suitable for self-calibration of these diverse types of cameras. In addition, many additional parameters might be highly correlated with interior orientation parameters or other correction parameters. A new generation of additional parameters based on Fourier series overcomes these shortcomings, but they are not suitable for analyzing and representing nonstationary distortion signals. This study develops a new model of wavelets-based additional parameters as well as its computation program system for self-calibrated bundle block adjustment. Preliminary tests are done by using aerial images in a calibration field. Their computation results demonstrate that wavelet additional parameters are helpful for correcting camera lens distortion. More details are shown in this paper.