Fringe filtering technique based on local signal reconstruction using noise subspace inflation. Issue 6 (25th March 2016)

• Record Type: Journal Article
• Authors: Kulkarni, Rishikesh; Rastogi, Pramod
• Is Part Of: Journal of modern optics. Volume 63:Issue 6(2016)
• Page Start: 545
• Page End: 552
• Publication Date: 2016-03-25
• Subjects: fringe filtering -- phase estimation -- matrix enhancement -- noise subspace inflation; Optics -- Periodicals; 535
• DOI: 10.1080/09500340.2015.1084055
• Languages: English
• ISSNs: 0950-0340

Abstract

Abstract : A noise filtering technique is proposed to filter the fringe pattern recorded in the optical measurement set-up. A single fringe pattern carrying the information on the measurand is treated as a data matrix which can either be complex or real valued. In the first approach, the noise filtering is performed pixel-wise in a windowed data segment generated around each pixel. The singular value decomposition of an enhanced form of this data segment is performed to extract the signal component from a noisy background. This enhancement of matrix has an effect of noise subspace inflation which accommodates maximum amount of noise. In another computationally efficient approach, the data matrix is divided into number of small-sized blocks and filtering is performed block-wise based on the similar noise subspace inflation method. The proposed method has an important ability to identify the spatially varying fringe density and regions of phase discontinuities. The performance of the proposed method is validated with numerical and experimental results.