Image De-Quantization Using Plate Bending Model

Discretized image signals might have a lower dynamic range than the display. Because of this, false contours might appear when the image has the same pixel value for a larger region and the distance between pixel levels reaches the noticeable difference threshold. There have been several methods aimed at approximating the high bit depth of the original signal. Our method models a region with a bended plate model, which leads to the biharmonic equation. This method addresses several new aspects: the reconstruction of non-continuous regions when foreground objects split the area into separate regions; the incorporation of confidence about pixel levels, making the model tunable; and the method gives a physics-inspired way to handle local maximal/minimal regions. The solution of the biharmonic equation yields a smooth high-order signal approximation and handles the local maxima/minima problems.

Efficient N-Dimensional surface estimation using Crofton formula and run-length encoding

Unlike the measure of the area in 2D or of the volume in 3D, the perimeter and the surface are not easily measurable in a discretized image. In this article we describe a method based on the Crofton formula to measure those two parameters in a discritized image. The accuracy of the method is discussed and tested on several known objects. An algorithm based on the run-length encoding of binary objects is presented and compared to other approaches. An implementation is provided and integrated in the LabelObject/LabelMap framework contributed earlier by the authors.

LABELING OF N-DIMENSIONAL IMAGES WITH CHOOSABLE ADJACENCY OF THE PIXELS

The labeling of discretized image data is one of the most essential operations in digital image processing. The notions of an adjacency system of pixels and the complementarity of two such systems are crucial to guarantee consistency of any labeling routine. In to date's publications, this complementarity usually is defined using discrete versions of the Jordan-Veblen curve theorem and the Jordan-Brouwer surface theorem for 2D and 3D images, respectively. In contrast, we follow here an alternative concept, which relies on a consistency relation for the Euler number. This relation and all necessary definitions are easily stated in a uniform manner for the n-dimensional case. For this, we present identification and convergence results for complementary adjacency systems, supplemented by examples for the 3D case. Next, we develop a pseudo-code for a general labeling algorithm. The application of such an algorithm should be assessed with regard to our preceding considerations. A benchmark and a thorough discussion finish our article.