Bootstrapping de-shadowing and self-calibration for scanning electron microscope photometric stereo
Atsushi Miyamoto, Deshan Chen, Shun’ichi Kaneko · Measurement Science and Technology · 2014
In this paper, we present a novel approach that addresses the blind reconstruction problem in scanning electron microscope (SEM) photometric stereo. Using only two observed images that suffer from shadowing effects, our method automatically calibrates the parameter and resolves shadowing errors for estimating an accurate three-dimensional (3D) shape and underlying shadowless images. We introduce a novel shadowing compensation model using image intensities for both cases of presence and absence of shadowing. With this model, the proposed de-shadowing algorithm iteratively compensates for image intensities and modifies the corresponding 3D surface. Besides de-shadowing, we introduce a practically useful self-calibration criterion by enforcing a good reconstruction. We show that incorrect parameters will engender significant distortions of 3D reconstructions in shadowed regions during the de-shadowing procedure. This motivated us to design the self-calibration criterion by utilizing shadowing to pursue the proper parameter that produces the best reconstruction with least distortions. As a result, we develop a bootstrapping approach for simultaneous de-shadowing and self-calibration in SEM photometric stereo. Extensive experiments on real image data demonstrate the effectiveness of our method.