Several methods have been proposed for the task of writer identification for historical manuscripts. Most of these methods have been evaluated on private historical datasets only, while few have been evaluated on the recently published historical dataset provided by the Historical-WI competition at ICDAR-2017. Nevertheless, there is no thorough analysis in the literature available w.r.t. the degradation typically found in the digitized manuscripts. Furthermore, the currently proposed methods are beyond the reach of the scholars from the humanities; either because of the impracticality of the method itself, or because of the lack of an easy-to-use implementation. In this paper, we analyse a state-of-the-art method against common degradation types in historical manuscripts using images from the virtual manuscript library of Switzerland. Furthermore, we show that, by optimising a key parameter, we can enhance the performance of the method and significantly outperform the winner method of the Historical-WI competition. Finally, we demonstrate the practicality of our implementation yielding intuitively comprehensible results for direct use of scholars from the humanities.

Writer identification and verification can be viewed as a classification problem, where each writer represents a class. We propose a classifier for offline, text-independent, and segmentation-free writer identification based on the Local Naïve Bayes Nearest-Neighbour (Local NBNN) classification. Our proposed method takes into consideration the particularity of handwriting patterns by adding a constraint to prevent the matching of irrelevant keypoints. Furthermore, a normalisation factor is proposed to cope with the prevalent problem of unbalanced data. The method has been evaluated on several public datasets of different writing systems and state-of-the-art results are shown to be improved.

We present a novel approach to the coarse segmentation of tubular structures in three-dimensional (3-D) image data. Our algorithm, which requires only few initial values and minimal user interaction, can be used to initialize complex deformable models and is based on an extension of the randomized hough transform (RHT), a robust method for low-dimensional parametric object detection. Tubular structures are modeled as generalized cylinders. By means of a discrete Kalman filter, they are tracked through 3-D space. Our extensions to the RHT are a feature adaptive selection of the sample size, expectation-dependent weighting of the input data, and a novel 3-D parameterization for straight elliptical cylinders. Experimental results obtained for 3-D synthetic as well as for 3-D medical images demonstrate the robustness of our approach w.r.t. image noise. We present the successful segmentation of tubular anatomical structures such as the aortic arc and the spinal cord.

We investigate several iterative numerical schemes for nonlinear variational image smoothing and segmentation implemented in parallel. A general iterative framework subsuming these schemes is suggested for which global convergence irrespective of the starting point can be shown. We characterize various edge-preserving regularization methods from the image processing literature involving auxiliary variables as special cases of this general framework. As a by-product, global convergence can be proven under conditions slightly weaker than these stated in the literature. Efficient Krylov subspace solvers for the linear parts of these schemes have been implemented on a multiprocessor machine. The performance of these parallel implementations has been assessed and empirical results concerning convergence rates and speed-up factors are reported.

We consider elastic image registration based on a set of corresponding anatomical point landmarks and approximating thin-plate splines. This approach is an extension of the original interpolating thin-plate spline approach and allows to take into account landmark localization errors. The extension is important for clinical applications since landmark extraction is always prone to error. Our approach is based on a minimizing functional and can cope with isotropic as well as anisotropic landmark errors. In particular, in the latter case it is possible to include different types of landmarks, e.g., unique point landmarks as well as arbitrary edge points. Also, the scheme is general with respect to the image dimension and the order of smoothness of the underlying functional. Optimal affine transformations as well as interpolating thin-plate splines are special cases of this scheme. To localize landmarks we use a semi-automatic approach which is based on three-dimensional (3-D) differential operators. Experimental results are presented for two-dimensional as well as 3-D tomographic images of the human brain.

We introduce radial basis functions with compact support for elastic registration of medical images. With these basis functions the influence of a landmark on the registration result is limited to a circle in 2D and, respectively, to a sphere in 3D. Therefore, the registration can be locally constrained which especially allows to deal with rather local changes in medical images due to, e.g., tumor resection. An important property of the used RBFs is that they are positive definite. Thus, the solvability of the resulting system of equations is always guaranteed. We demonstrate our approach for synthetic as well as for 2D and 3D tomographic images.

The accuracy of image-guided neurosurgery generally suffers from brain deformations due to intraoperative changes. These deformations cause significant changes of the anatomical geometry (organ shape and spatial interorgan relations), thus making intraoperative navigation based on preoperative images error prone. In order to improve the navigation accuracy, we developed a biomechanical model of the human head based on the finite element method, which can be employed for the correction of preoperative images to cope with the deformations occurring during surgical interventions. At the current stage of development, the two-dimensional (2-D) implementation of the model comprises two different materials, though the theory holds for the three-dimensional (3-D) case and is capable of dealing with an arbitrary number of different materials. For the correction of a preoperative image, a set of homologous landmarks must be specified which determine correspondences. These correspondences can be easily integrated into the model and are maintained throughout the computation of the deformation of the preoperative image. The necessary material parameter values have been determined through a comprehensive literature study. Our approach has been tested for the case of synthetic images and yields physically plausible deformation results. Additionally, we carried out registration experiments with a preoperative MR image of the human head and a corresponding postoperative image simulating an intraoperative image. We found that our approach yields good prediction results, even in the case when correspondences are given in a relatively small area of the image only.

The paper presentsanewparameter-free approach to non-rigid image registration, where displacements, obtained through a mapping of boundary structures in the source and target image, are incorporated as hard constraints for elastic image deformation. As a consequence, our approach does not contain any parameters of the deformation model (elastic constants). The approach guarantees the exact correspondence of boundary structures after elastic transformation provided that correct input data are available. We describe a linear and an incremental model, the latter model allows to cope also with large deformations. Experimental results for 2-D and 3-D synthetic as well as real medical images are presented. 1

A novel approach has been proposed for generating a corresponding pair of straight lines in biplane angiograms. Such lines are very useful in 3D analysis of biplane angiograms, such as a reconstruction of 3D arteries, a precise matching of arterial diameter measurements in biplane images, etc.>

Differently scaled intensity discontinuities along the gradient direction of scaled oriented contrast edges define significant X-ray CT and MR image structures. Reliable detection and quantitative description of such scaled intensity discontinuities is achieved with a scale-space approach. It results in an explicit token representation which can be understood as a compact symbolic description of significant image structure, namely a set of quantitative attributes per contrast edge. Moreover, a multi-resolution approach is used to reconstruct the image intensity function from the token representation alone using a membrane as prior image model. Current results from the novel entire processing cascade, which combines a quantitative discontinuity description and a membrane-based intensity reconstruction, are presented.>