An effective assessment method for absolute phase retrieval in digital fringe projection profilometry

Abstract

In this paper, we propose an effective method for assessing absolute phase retrieval quality in digital fringe projection profilometry (DFPP). The method provides a quantitative analysis of the phase retrieval quality with a new integrated quality metric. The phase map of the measured object is pre-defined on the imaging plane and called the inverse ground truth. To obtain the measurement result of the inverse ground truth, object-adapted patterns are projected onto the object and the phase map of the object is retrieved by a phase retrieval technique. Subsequently, the retrieved phase map of the object is compared with the inverse ground truth to assess the phase retrieval quality. The structural similarity between the measurement result and the inverse ground truth is quantified by the proposed quality metric in terms of mean, contrast and structure comparisons. Experiments are carried out to evaluate the effectiveness of the proposed method from two aspects: comparing the accuracies of different phase retrieval techniques and evaluating the influence of different setup parameters on the accuracy. Experimental results show that the analysis by the proposed quality metric is well consistent with the existing findings for the phase retrieval quality in DFPP.