Monocular Microscopic Image 3D Reconstruction Algorithm based on Depth from Defocus with Adaptive Window Selection

At present, in the field of microscopic measurement, 3D reconstruction of microscopic objects is still a challenging topic. Considering the characteristics of microscopy imaging, such as high complexity and the limited depth of field, this paper applies a 3D reconstruction algorithm based on the depth from defocus (DFD) technique. In DFD, focus measure plays an important role. However, there is a limit in traditional focus measure, which is that a fixed window has been used at each stage of focus measure in DFD. A smaller window is unable to cover enough field information in the case of smoothness or less texture. Whereas, a lager window may over-smooth or distort the shape of object. Based on these factors, without using a fixed window, this paper proposed a scheme selecting adaptive windows for 3D reconstruction. Comprehensive verification experiments with image sequences of real objects have reported the effectiveness of the proposed algorithm.

[1]  Shree K. Nayar,et al.  Shape from Focus , 1994, IEEE Trans. Pattern Anal. Mach. Intell..

[2]  Yujuan Sun,et al.  Reformed Residual Network With Sparse Feedbacks for 3D Reconstruction From a Single Image , 2018, IEEE Access.

[3]  Zhiming Xu,et al.  Three-dimensional surface microtopography recovery from a multifocus image sequence using an omnidirectional modified Laplacian operator with adaptive window size. , 2017, Applied optics.

[4]  Shutao Li,et al.  Pixel-level image fusion: A survey of the state of the art , 2017, Inf. Fusion.

[5]  Shree K. Nayar,et al.  Real-time focus range sensor , 1995, Proceedings of IEEE International Conference on Computer Vision.

[6]  R. Leach.  Optical measurement of surface topography , 2011 .

[7]  Adrien Bartoli,et al.  A Stable Analytical Framework for Isometric Shape-from-Template by Surface Integration , 2017, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[8]  Tae-Sun Choi,et al.  Improving focus measurement via variable window shape on surface radiance distribution for 3D shape reconstruction , 2013 .

[9]  Takaaki Akimoto,et al.  Automatic creation of 3D facial models , 1993, IEEE Computer Graphics and Applications.

[10]  Francesc Moreno-Noguer,et al.  A Bayesian approach to simultaneously recover camera pose and non-rigid shape from monocular images , 2016, Image Vis. Comput..

[11]  Jay Martin Tenenbaum,et al.  Accommodation in computer vision , 1971 .

[12]  Tae-Sun Choi,et al.  Adaptive window selection for 3D shape recovery from image focus , 2013 .

[13]  Noah Snavely,et al.  Scene Reconstruction and Visualization from Internet Photo Collections: A Survey , 2011, IPSJ Trans. Comput. Vis. Appl..

[14]  Aamir Saeed Malik,et al.  Real-time processing for shape-from-focus techniques , 2012, Journal of Real-Time Image Processing.

[15]  Song Bai,et al.  Deep learning representation using autoencoder for 3D shape retrieval , 2014, Proceedings 2014 IEEE International Conference on Security, Pattern Analysis, and Cybernetics (SPAC).