Passive navigation as a pattern recognition problem

The most basic visual capabilities found in living organisms are based on motion. Machine vision, of course, does not have to copy animal vision, but the existence of reliably functioning vision modules in nature gives us some reason to believe that it is possible for an artificial system to work in the same or a similar way. In this article it is argued that many navigational capabilities can be formulated as pattern recognition problems. An appropriate retinotopic representation of the image would make it possible to extract the information necessary to solve motion-related tasks through the recognition of a set of locations on the retina. This argument is illustrated by introducing a representation of image motion by which an observer's egomotion could be derived from information globally encoded in the image-motion field. In the past, the problem of determining a system's own motion from dynamic imagery has been considered as one of the classical visual reconstruction problems, wherein local constraints have been employed to compute from exact 2-D image measurements (correspondence, optical flow) the relative 3-D motion and structure of the scene in view. The approach introduced here is based on new global constraints defined on local normal-flow measurements—the spatio-temporal derivatives of the image-intensity function. Classifications are based on orientations of normal-flow vectors, which allows selection of vectors that form global patterns in the image plane. The position of these patterns is related to the 3-D motion of the observer, and their localization provides the axis of rotation and the direction of translation. The constraints introduced are utilized in algorithmic procedures formulated as search techniques. These procedures are very stable, since they are not affected by small perturbations in the image measurements. As a matter of fact, the solution to the two directions of translation and rotation is not affected, as long as the measurement of the sign of the normal flow is correct.Part of this work was accomplished while the author was visiting the Royal Institute of Technology, Stockholm, Sweden. The research was supported in part by ARPA, ONR, and Swedish National Science Foundation.     

Observability of 3D Motion

This paper examines the inherent difficulties in observing 3D rigid motion from image sequences. It does so without considering a particular estimator. Instead, it presents a statistical analysis of all the possible computational models which can be used for estimating 3D motion from an image sequence. These computational models are classified according to the mathematical constraints that they employ and the characteristics of the imaging sensor (restricted field of view and full field of view). Regarding the mathematical constraints, there exist two principles relating a sequence of images taken by a moving camera. One is the epipolar constraint, applied to motion fields, and the other the positive depth constraint, applied to normal flow fields. 3D motion estimation amounts to optimizing these constraints over the image. A statistical modeling of these constraints leads to functions which are studied with regard to their topographic structure, specifically as regards the errors in the 3D motion parameters at the places representing the minima of the functions. For conventional video cameras possessing a restricted field of view, the analysis shows that for algorithms in both classes which estimate all motion parameters simultaneously, the obtained solution has an error such that the projections of the translational and rotational errors on the image plane are perpendicular to each other. Furthermore, the estimated projection of the translation on the image lies on a line through the origin and the projection of the real translation. The situation is different for a camera with a full (360 degree) field of view (achieved by a panoramic sensor or by a system of conventional cameras). In this case, at the locations of the minima of the above two functions, either the translational or the rotational error becomes zero, while in the case of a restricted field of view both errors are non-zero. Although some ambiguities still remain in the full field of view case, the implication is that visual navigation tasks, such as visual servoing, involving 3D motion estimation are easier to solve by employing panoramic vision. Also, the analysis makes it possible to compare properties of algorithms that first estimate the translation and on the basis of the translational result estimate the rotation, algorithms that do the opposite, and algorithms that estimate all motion parameters simultaneously, thus providing a sound framework for the observability of 3D motion. Finally, the introduced framework points to new avenues for studying the stability of image-based servoing schemes.     

Structure from Motion: Beyond the Epipolar Constraint

The classic approach to structure from motion entails a clear separation between motion estimation and structure estimation and between two-dimensional (2D) and three-dimensional (3D) information. For the recovery of the rigid transformation between different views only 2D image measurements are used. To have available enough information, most existing techniques are based on the intermediate computation of optical flow which, however, poses a problem at the locations of depth discontinuities. If we knew where depth discontinuities were, we could (using a multitude of approaches based on smoothness constraints) accurately estimate flow values for image patches corresponding to smooth scene patches; but to know the discontinuities requires solving the structure from motion problem first. This paper introduces a novel approach to structure from motion which addresses the processes of smoothing, 3D motion and structure estimation in a synergistic manner. It provides an algorithm for estimating the transformation between two views obtained by either a calibrated or uncalibrated camera. The results of the estimation are then utilized to perform a reconstruction of the scene from a short sequence of images.The technique is based on constraints on image derivatives which involve the 3D motion and shape of the scene, leading to a geometric and statistical estimation problem. The interaction between 3D motion and shape allows us to estimate the 3D motion while at the same time segmenting the scene. If we use a wrong 3D motion estimate to compute depth, we obtain a distorted version of the depth function. The distortion, however, is such that the worse the motion estimate, the more likely we are to obtain depth estimates that vary locally more than the correct ones. Since local variability of depth is due either to the existence of a discontinuity or to a wrong 3D motion estimate, being able to differentiate between these two cases provides the correct motion, which yields the least varying estimated depth as well as the image locations of scene discontinuities. We analyze the new constraints, show their relationship to the minimization of the epipolar constraint, and present experimental results using real image sequences that indicate the robustness of the method.     

Ambiguity in Structure from Motion: Sphere versus Plane

If 3D rigid motion can be correctly estimated from image sequences, the structure of the scene can be correctly derived using the equations for image formation. However, an error in the estimation of 3D motion will result in the computation of a distorted version of the scene structure. Of computational interest are these regions in space where the distortions are such that the depths become negative, because in order for the scene to be visible it has to lie in front of the image, and thus the corresponding depth estimates have to be positive. The stability analysis for the structure from motion problem presented in this paper investigates the optimal relationship between the errors in the estimated translational and rotational parameters of a rigid motion that results in the estimation of a minimum number of negative depth values. The input used is the value of the flow along some direction, which is more general than optic flow or correspondence. For a planar retina it is shown that the optimal configuration is achieved when the projections of the translational and rotational errors on the image plane are perpendicular. Furthermore, the projection of the actual and the estimated translation lie on a line through the center. For a spherical retina, given a rotational error, the optimal translation is the correct one; given a translational error, the optimal rotational negative deptherror depends both in direction and value on the actual and estimated translation as well as the scene in view. The proofs, besides illuminating the confounding of translation and rotation in structure from motion, have an important application to ecological optics. The same analysis provides a computational explanation of why it is easier to estimate self-motion in the case of a spherical retina and why shape can be estimated easily in the case of a planar retina, thus suggesting that nature's design of compound eyes (or panoramic vision) for flying systems and camera-type eyes for primates (and other systems that perform manipulation) is optimal.     

‘Long autonomy or long delay?’ The importance of domain in opinion mining

Nowadays, people do not only navigate the web, but they also contribute contents to the Internet. Among other things, they write their thoughts and opinions in review sites, forums, social networks, blogs and other websites. These opinions constitute a valuable resource for businesses, governments and consumers. In the last years, some researchers have proposed opinion extraction systems, mostly domain-independent ones, to automatically extract structured representations of opinions contained in those texts. In this work, we tackle this task in a domain-oriented approach, defining a set of domain-specific resources which capture valuable knowledge about how people express opinions on a given domain. These resources are automatically induced from a set of annotated documents. Some experiments were carried out on three different domains (user-generated reviews of headphones, hotels and cars), comparing our approach to other state-of-the-art, domain-independent techniques. The results confirm the importance of the domain in order to build accurate opinion extraction systems. Some experiments on the influence of the dataset size and an example of aggregation and visualization of the extracted opinions are also shown.     

Ultra‐Low‐Loss Acrylate Polymers for Planar Light Circuits

We have developed a materials system composed of liquid multifunctional fluorinated acrylate monomers that achieves very low absorption losses within both datacom and telecom wavelength ranges. Identified structure–property relationships have guided the design of polymers that are optimized with respect to their inherent materials properties and their ability to be processed into low‐loss optical waveguides. Together these co‐developed materials and processes combine to produce waveguides that have propagation losses equivalent to planar glass guides, and significantly reduced polarization dependence and improved thermal response needed for the performance of thermo‐optic devices. Extensive environmental studies have demonstrated robustness well beyond that required for normal operating conditions.     

Disposable receptor-based optical sensor for nitrate

A new optical absorption-based disposable sensor for nitrate is described. The nitrate-sensitive element is a bicyclic cyclophane receptor next to a suitable pH-sensitive lipophilic dye immobilized in a plasticized polymeric membrane. The rigid amide-based receptor with C3 symmetry controls the anion selectivity pattern of the optical element. The optical selectivity coefficients obtained for nitrate over a variety of naturally occurring anions in natural waters meet the requirements for the determination of nitrate in waters. The disposable sensor responds to nitrate rapidly-the typical response time is 5 min-and reversibly over a wide dynamic range (26 microM-63 mM) with sensor-to-sensor reproducibility (relative standard deviation, RSD, 3.68%, as log aNO3-, at the medium level of the range and RSD 1.39% for repeated measurements with the same sensor). The performance of the optical disposable sensor was tested for the analysis of nitrate in different types of natural waters (river, well, spring), validating results against a reference procedure. The proposed method is quick, inexpensive, selective, and sensitive and uses only conventional instrumentation.     

Engineering and manufacturing of ITER first mirror mock-ups

Most of the ITER optical diagnostics aiming at viewing and monitoring plasma facing components will use in-vessel metallic mirrors. These mirrors will be exposed to a severe plasma environment and lead to an important tradeoff on their design and manufacturing. As a consequence, investigations are carried out on diagnostic mirrors toward the development of optimal and reliable solutions. The goals are to assess the manufacturing feasibility of the mirror coatings, evaluate the manufacturing capability and associated performances for the mirrors cooling and polishing, and finally determine the costs and delivery time of the first prototypes with a diameter of 200 and 500 mm. Three kinds of ITER candidate mock-ups are being designed and manufactured: rhodium films on stainless steel substrate, molybdenum on TZM substrate, and silver films on stainless steel substrate. The status of the project is presented in this paper.     

A statistical view of iterative methods for linear inverse problems

In this article, we study the problem of recovering the unknown solution of a linear ill-posed problem, via iterative regularization methods, from a statistical point of view. The basic purpose of the paper is to develop adaptive model selection techniques for determining the regularization parameters, i.e., the iteration index. We assume observations are taken over a fixed grid and we consider solutions over a sequence of finite-dimensional subspaces. Based on concentration inequalities techniques, we derive non-asymptotic optimal upper bounds for the mean square error of the proposed estimator.