In this paper, we propose a technique to perform image matching, which is invariant to changes in geometry and intensity. In detail, this approach relies on a representation of an image portion as a 3D function simultaneously coding both the 2D spatial coordinates and the intensity value of each pixel. In this equivalent image representation, a geometric and intensity variation between the visual content of two images depicting the same objects leads to an isometric relationship between the image 3D functions. The comparison between the two images is effectively obtained by estimating the 3D linear isometry that connects them. The effectiveness of the proposed technique was assessed on two synthetic, and one real, image data sets.
Intensity and Affine Invariant Statistic-based Image Matching
Lecca, Michela;
2012-01-01
Abstract
In this paper, we propose a technique to perform image matching, which is invariant to changes in geometry and intensity. In detail, this approach relies on a representation of an image portion as a 3D function simultaneously coding both the 2D spatial coordinates and the intensity value of each pixel. In this equivalent image representation, a geometric and intensity variation between the visual content of two images depicting the same objects leads to an isometric relationship between the image 3D functions. The comparison between the two images is effectively obtained by estimating the 3D linear isometry that connects them. The effectiveness of the proposed technique was assessed on two synthetic, and one real, image data sets.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
