Large carnivore habitat suitability modelling for Romania and associated predictions for protected areas.

Habitat characteristics associated with species occurrences represent important baseline information for wildlife management and conservation, but have rarely been assessed for countries recently joining the EU. We used footprint tracking data and landscape characteristics in Romania to investigate the occurrence of brown bear (Ursus arctos), gray wolf (Canis lupus) and Eurasian lynx (Lynx lynx) and to compare model predictions between Natura 2000 and national-level protected areas (gap analysis). Wolves were more likely to occur where rugged terrain was present. Increasing proportion of forest was positively associated with occurrence of all large carnivores, but forest type (broadleaf, mixed, or conifer) generally varied with carnivore species. Areas where cultivated lands were extensive had little suitable habitat for lynx, whereas bear occurrence probability decreased with increasing proportion of built areas. Pastures were positively associated with wolf and lynx occurrence. Brown bears occurred primarily where national roads with high traffic volumes were at low density, while bears and lynx occurred at medium-high densities of communal roads that had lower traffic volumes. Based on predictions of carnivore distributions, natural areas protected in national parks were most suitable for carnivores, nature parks were less suitable, whereas EU-legislated Natura 2000 sites had the lowest probability of carnivore presence. Our spatially explicit carnivore habitat suitability predictions can be used by managers to amend borders of existing sites, delineate new protected areas, and establish corridors for ecological connectivity. To assist recovery and recolonization, management could also focus on habitat predicted to be suitable but where carnivores were not tracked.

Realigning 2D and 3D Object Fragments without Correspondences.

This paper addresses the problem of simultaneous estimation of different linear deformations, resulting in a global non-linear transformation, between an original object and its broken fragments. A general framework is proposed without using correspondences, where the solution of a polynomial system of equations directly provides the parameters of the alignment. We quantitatively evaluate the proposed algorithm on a large synthetic dataset containing 2D and 3D images, where linear (rigid-body and affine) transformations are considered. We also conduct an exhaustive analysis of the robustness against segmentation errors and the numerical stability of the proposed method. Moreover, we present experiments on 2D real images as well as on volumetric medical images.

Nonlinear shape registration without correspondences.

In this paper, we propose a novel framework to estimate the parameters of a diffeomorphism that aligns a known shape and its distorted observation. Classical registration methods first establish correspondences between the shapes and then compute the transformation parameters from these landmarks. Herein, we trace back the problem to the solution of a system of nonlinear equations which directly gives the parameters of the aligning transformation. The proposed method provides a generic framework to recover any diffeomorphic deformation without established correspondences. It is easy to implement, not sensitive to the strength of the deformation, and robust against segmentation errors. The method has been applied to several commonly used transformation models. The performance of the proposed framework has been demonstrated on large synthetic data sets as well as in the context of various applications.