In this work, we extend a common framework for graph-based image segmentation that includes the graph cuts, random walker, and shortest path optimization algorithms. Viewing an image as a weighted graph, these algorithms can be expressed by means of a common energy function with differing choices of a parameter q acting as an exponent on the differences between neighboring nodes. Introducing a new parameter p that fixes a power for the edge weights allows us to also include the optimal spanning forest algorithm for watershed in this same framework. We then propose a new family of segmentation algorithms that fixes p to produce an optimal spanning forest but varies the power q beyond the usual watershed algorithm, which we term the power watershed. In particular, when q=2, the power watershed leads to a multilabel, scale and contrast invariant, unique global optimum obtained in practice in quasi-linear time. Placing the watershed algorithm in this energy minimization framework also opens new possibilities for using unary terms in traditional watershed segmentation and using watershed to optimize more general models of use in applications beyond image segmentation.
In this paper we address the computation of globally minimal curves and surfaces for image segmentation and stereo reconstruction. We present a solution, simulating a continuous maximal flow by a novel system of partial differential equations. Existing methods are either grid-biased (graph-based methods) or sub-optimal (active contours and surfaces).The solution simulates the flow of an ideal fluid with isotropic velocity constraints. Velocity constraints are defined by a metric derived from image data. An auxiliary potential function is introduced to create a system of partial differential equations. It is proven that the algorithm produces a globally maximal continuous flow at convergence, and that the globally minimal surface may be obtained trivially from the auxiliary potential. The bias of minimal surface methods toward small objects is also addressed.An efficient implementation is given for the flow simulation.The globally minimal surface algorithm is applied to segmentation in 2D and 3D as well as to stereo matching. Results in 2D agree with an existing minimal contour algorithm for planar images.Results in 3D segmentation and stereo matching demonstrate that the new algorithm is robust and free from grid bias.
In this work, we consider a class of differentiable criteria for sparse image computing problems, where a nonconvex regularization is applied to an arbitrary linear transform of the target image. As special cases, it includes edge preserving measures or frame-analysis potentials commonly used in image processing. As shown by our asymptotic results, the ℓ 2 − ℓ 0 penalties we consider may be employed to provide approximate solutions to ℓ 0penalized optimization problems. One of the advantages of the proposed approach is that it allows us to derive an efficient Majorize-Minimize subspace algorithm. The convergence of the algorithm is investigated by using recent results in nonconvex optimization. The fast convergence properties of the proposed optimization method are illustrated through image processing examples. In particular, its effectiveness is demonstrated on several data recovery problems. * A preliminary version of this work has appeared in .
International audienceVery high spatial resolution (VHR) images allow to feature man-made structures such as roads and thus enable their accurate analysis. Geometrical characteristics can be extracted using mathematical morphology. However, the prior choice of a reference shape (structuring element) introduces a shape-bias. This paper presents a new method for extracting roads in Very High Resolution remotely sensed images based on advanced directional morphological operators. The proposed approach introduces the use of Path Openings and Path Closings in order to extract structural pixel information. These morphological operators remain flexible enough to fit rectilinear and slightly curved structures since they do not depend on the choice of a structural element shape. As a consequence, they outperform standard approaches using rotating rectangular structuring elements. The method consists in building a granulometry chain using Path Openings and Path Closing to construct Morphological Profiles. For each pixel, the Morphological Profile constitutes the feature vector on which our road extraction is based
The SARS-COV-2 pandemic has put pressure on intensive care units, so that identifying predictors of disease severity is a priority. We collect 58 clinical and biological variables, and chest CT scan data, from 1003 coronavirus-infected patients from two French hospitals. We train a deep learning model based on CT scans to predict severity. We then construct the multimodal AI-severity score that includes 5 clinical and biological variables (age, sex, oxygenation, urea, platelet) in addition to the deep learning model. We show that neural network analysis of CT-scans brings unique prognosis information, although it is correlated with other markers of severity (oxygenation, LDH, and CRP) explaining the measurable but limited 0.03 increase of AUC obtained when adding CT-scan information to clinical variables. Here, we show that when comparing AI-severity with 11 existing severity scores, we find significantly improved prognosis performance; AI-severity can therefore rapidly become a reference scoring approach.
Adiposity is associated with chronic low-grade systemic inflammation and increased inflammation in the hypothalamus, a key structure in feeding behavior. It remains unknown whether inflammation impacts other brain structures that regulate feeding behavior. We studied 44 overweight/obese and 19 lean individuals with MRI and plasma fibrinogen levels (marker of inflammation). We performed MRI-based segmentations of the medial and lateral orbitofrotal cortex (OFC) and hippocampal volumes. Gray matter (GM) volumes were adjusted for head size variability. We conducted logistic and hierarchical regressions to assess the association between fibrinogen levels and brain volumetric data. Using diffusion tensor imaging (DTI), we created apparent diffusion coefficient (ADC) maps and conducted voxelwise correlational analyses. Fibrinogen concentrations were higher among the overweight/obese (t=-2.33, P=0.023). Lateral OFC associated together with fibrinogen correctly classified those with excess of weight (accuracy=76.2%, sensitivity=95.5%, specificity=31.6%). The lateral OFC volumes of overweight/obese were negatively associated with fibrinogen (r=-0.37, P=0.016) and after accounting for age, hypertension, waist/hip ratio and lipid and sugar levels, fibrinogen significantly explained an additional 9% of the variance in the lateral OFC volume (β=-0.348, ΔR2=0.093, ΔF P=0.046). Among overweight/obese the associations between GM ADC and fibrinogen were significantly positive (P<0.001) in the left and right amygdala and the right parietal region. Among lean individuals these associations were negative and located in the left prefrontal, the right parietal and the left occipital lobes. This is the first study to report that adiposity-related inflammation may reduce the integrity of some of the brain structures involved in reward and feeding behaviors.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
334 Leonard St
Brooklyn, NY 11211
Copyright © 2023 scite Inc. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.