New Holistic Self-Calibration Method for Deflectometric Sensors

Phase Measuring Deflectometry (PMD) is a favorable method to precisely measure the local slope of specular surfaces. To acquire the global shape, we already presented methods to integrate the slope data. However, integration largely amplifies low frequency systematic errors. As a consequence, a global shape accuracy better than 1µm on a field of 80mm requires an extremely accurate calibration of the sensor. Such an accurate calibration can hardly be achieved by common methods. The standard approach includes several isolated steps to calibrate all sensor components individually. This commonly yields inconsistent results. We will present a novel calibration method that needs only measurements using a planar mirror in several arbitrary positions. Based on this data, all calibration parameters for the system are estimated in one single optimization process. The new approach has a number of significant advantages: It accomplishes a more accurate and stable calibration. The concept is open to improve the accuracy even further by increasing the number of cameras. As an additional benefit for the operator in the field, it neither requires special calibration targets nor any user interaction.