Frédérique Bicking

Estimation of depth on thick edges from sharp and blurred images

This article deals with the generalization of a local depth estimation method using sharp edges and blurred edges. This Depth from Defocus method is explained and the theoretical relations are defined. Improvements concerning the generalization and the noise sensitivity on the depth estimation are developed and application conditions are exposed. Some results on synthetic images are presented to illustrate the method efficiency.

Hybrid computation of uncertainty in reliability analysis with p-box and evidential networks

This paper presents a method to assess system reliability in the presence of epistemic and aleatory uncertainty. This hybrid method uses belief functions to model and manipulate uncertainty. P-boxes are used to represent basic uncertainties and acyclic directed networks to model the system reliability. These choices allow a flexible modeling of uncertainty, while limiting the computational cost of inferences. In particular, they offer convenient ways of integrating expert opinions and many kinds of uncertainty sources. It also can model complex systems. In this paper, we introduce the modeling method and apply it to a fire detection system with different kinds of input p-boxes for the sake of illustration.

Integration of dependability in a task allocation problem

The task allocation on various components or processing units composing an automation system needs a design-aided methodology in order to obtain both relevant architecture of this automation system and a sufficiently efficient task allocation. The dependability, mainly illustrated in our approach by the availability of the process and by the reliability that characterizes the risk of incidents or accident occurrences, can be taken into account in the earliest stage of the automation system design. The main interest is that it allows for the definition of reliable operational and functional architectures. To achieve the task allocation, the search method used must integrate a notion of cost that includes both technological and functional aspects and those linked to the dependability. A genetic approach is used for the task allocation, giving a very satisfactory result in a reasonable computational time.

Control Allocation Using Reliability Measures for Over-Actuated System

An optimal control law synthesis conditioned by the reliability importance measures and degradation of components in the presence of failures is presented. The aim is to preserve the health of the actuators and the availability of over-actuated system both in the nominal situation and in the presence of some actuator failures. The reliability importance measures are computed by mean of a Bayesian Network (BN) and degradation is modeled by a Dynamic Bayesian Network (DBN) since BN and DBN are well suited to compute actuators reliability through the bayesian inference. It is applied on a system to estimate its reliability and those of its components and provide the parameters to synthesize the control laws. The performance of the developed method is illustrated on an hydraulic system.

Depth estimation based on thick oriented edges in images

The field of contact free 3D measure is particularly open since current measuring devices use lines or points. In this domain, cameras are the preferential sensors to develop global 3D measurement system. Many methods have been developed to obtain the 3D coordinates of objects using images and all exploit the variations of acquisition parameters. Acquisition parameters of the system or of the luminous environment, controlled or not are the essential information to establish a relationship between the image and the real scene. This article deals with a spatial approach to depth estimation by analysis of edges in images. A depth from defocus method is explained and the physical process is described. Theoretical developments are made to apply it on thin or thick edges and for height orientations of the edge profile. Some results on images are presented to illustrate the efficiency and the influence of the conditions of use.

Influence of image processing operators and optical models used on the quality of depth estimation

Absfrnd This able deals with a spoficll approoch Io depfh csfimntion by analysis o/edges in imega. A Depthfrom Defooeus methodis exploinednnd thephysicalprocess is described Theoretid developments w e made to op ply if on fhin or thick edges and fhc imflaence o/fhc used imageprocessing operators and opticd models ispointed DUI and discussed Some results 01 images are presented Io illusbate the mefhod efleieney and the oppropriafrness of fhc used modeis in noisy cmf&