A fast pixel matching method based on phase feature extraction in online phase-measuring profilometry. Issue 18 (11th October 2017)

• Record Type: Journal Article
• Authors: Xu, Xingfen; Cao, Yiping; Wang, Yapin; Chen, Cheng; Fu, Guangkai; Sun, Song
• Is Part Of: Journal of modern optics. Volume 64:Issue 18(2017)
• Page Start: 1907
• Page End: 1914
• Publication Date: 2017-10-11
• Subjects: Information optics -- pixel matching -- phase feature extraction -- normalization and binarization -- the equivalent phase-shifting algorithm; Optics -- Periodicals; 535
• DOI: 10.1080/09500340.2017.1326637
• Languages: English
• ISSNs: 0950-0340

Abstract

Abstract: A fast pixel matching method based on phase feature extraction in online phase-measuring profilometry is proposed, which is used to solve the problem where the object motion is along a straight line. The movement of the object results in the displacement of images in the captured deformed patterns, so pixel matching is carried out to modify the positions of images, which not only realizes the point-to-point correspondence of the object, but extracts the equivalent phase-shifting deformed patterns. So, the 3D surface shape of the object moving on the pipeline can be reconstructed successfully with an equivalent phase-shifting algorithm. The equiphase surface phase feature extracted from the phase distributions of the object obtained by Fourier transform profilometry is used to assist pixel matching in this paper. Normalization and further binarization of the phase feature extraction improve pixel matching's accuracy and speed. The simulation and experiments verify the feasibility and validity of the proposed method.