Multi-sensor data fusion for measurement of complex freeform surfaces

Along with the rapid development of the science and technology in fields such as space optics, multi-scale enriched freeform surfaces are widely used to enhance the performance of the optical systems in both functionality and size reduction. Multi-sensor technology is considered as one of the promising methods to measure and characterize these surfaces at multiple scales. This paper presents a multi-sensor data fusion based measurement method to purposely extract the geometric information of the components with different scales which is used to establish a holistic geometry of the surface via data fusion. To address the key problems of multi-sensor data fusion, an intrinsic feature pattern based surface registration method is developed to transform the measured datasets to a common coordinate frame. Gaussian zero-order regression filter is then used to separate each measured data in different scales, and the datasets are fused based on an edge intensity data fusion algorithm within the same wavelength. The fused data at different scales is then merged to form a new surface with holistic multiscale information. Experimental study is presented to verify the effectiveness of the proposed method.