Introducing shape constraints into object-based traveltime tomography

• Published in: Inverse problems Vol. 32; no. 9; pp. 95002 - 95029
• Main Authors: Gaullier, G, Charbonnier, P, Heitz, F, Côte, P
• Format: Journal Article
• Language: English
• Published: IOP Publishing 01.09.2016
• Subjects: Engineering Sciences; geoacoustic inversion; level set method; shape-constrained reconstruction; Signal and Image processing; traveltime tomography
• ISSN: 0266-5611; 1361-6420
• DOI: 10.1088/0266-5611/32/9/095002

Traveltime tomography is a difficult, ill-posed reconstruction problem due to the nonlinearity of the forward model and the limited number of measurements usually available. In such an adverse situation, pixel-based regularization methods are generally unable to provide satisfactory reconstructions. In this paper we propose a novel object-based reconstruction method that introduces prior information about the shape of the structures to be reconstructed, which yields high quality geoacoustic inversion. The proposed method approaches the forward model by a series of linear problems, leading to a sequence of minimizations during which the shape prior is introduced. The method is demonstrated on synthetic and real data, collected on a specific bench dedicated to non-destructive testing of civil engineering structures.