Layered Graphical Models for Tracking Partially-Occluded Objects

Partial occlusions are commonplace in a variety of real world computer vision applications: surveillance, intelligent environments, assistive robotics, autonomous navigation, etc. While occlusion handling methods have been proposed, most methods tend to break down when confronted with numerous occluders in a scene. In this paper, a layered image-plane representation for tracking people through substantial occlusions is proposed. An image-plane representation of motion around an object is associated with a pre-computed graphical model, which can be instantiated efficiently during online tracking. A global state and observation space is obtained by linking transitions between layers. A Reversible Jump Markov Chain Monte Carlo approach is used to infer the number of people and track them online. The method outperforms two state-of-the-art methods for tracking over extended occlusions, given videos of a parking lot with numerous vehicles and a laboratory with many desks and workstations.

Tracking a Large Number of Objects from Multiple Views

We propose a multi-object multi-camera framework for tracking large numbers of tightly-spaced objects that rapidly move in three dimensions. We formulate the problem of finding correspondences across multiple views as a multidimensional assignment problem and use a greedy randomized adaptive search procedure to solve this NP-hard problem efficiently. To account for occlusions, we relax the one-to-one constraint that one measurement corresponds to one object and iteratively solve the relaxed assignment problem. After correspondences are established, object trajectories are estimated by stereoscopic reconstruction using an epipolar-neighborhood search. We embedded our method into a tracker-to-tracker multi-view fusion system that not only obtains the three-dimensional trajectories of closely-moving objects but also accurately settles track uncertainties that could not be resolved from single views due to occlusion. We conducted experiments to validate our greedy assignment procedure and our technique to recover from occlusions. We successfully track hundreds of flying bats and provide an analysis of their group behavior based on 150 reconstructed 3D trajectories.

3D Trajectory Recovery for Tracking Multiple Objects and Trajectory Guided Recognition of Actions

A mechanism is proposed that integrates low-level (image processing), mid-level (recursive 3D trajectory estimation), and high-level (action recognition) processes. It is assumed that the system observes multiple moving objects via a single, uncalibrated video camera. A novel extended Kalman filter formulation is used in estimating the relative 3D motion trajectories up to a scale factor. The recursive estimation process provides a prediction and error measure that is exploited in higher-level stages of action recognition. Conversely, higher-level mechanisms provide feedback that allows the system to reliably segment and maintain the tracking of moving objects before, during, and after occlusion. The 3D trajectory, occlusion, and segmentation information are utilized in extracting stabilized views of the moving object. Trajectory-guided recognition (TGR) is proposed as a new and efficient method for adaptive classification of action. The TGR approach is demonstrated using "motion history images" that are then recognized via a mixture of Gaussian classifier. The system was tested in recognizing various dynamic human outdoor activities; e.g., running, walking, roller blading, and cycling. Experiments with synthetic data sets are used to evaluate stability of the trajectory estimator with respect to noise.

Cachability of Web Objects

Much work on the performance of Web proxy caching has focused on high-level metrics such as hit rate and byte hit rate, but has ignored all the information related to the cachability of Web objects. Uncachable objects include those fetched by dynamic requests, objects with uncachable HTTP status code, objects with the uncachable HTTP header, objects with an HTTP 1.0 cookie, and objects without a last-modified header. Although some researchers filter the Web traces before they use them for analysis or simulation,many do not have a comprehensive understanding of the cachability of Web objects. In this paper we evaluate all the reasons that a Web object might be uncachable. We use traces from NLANR. Since these traces do not contain HTTP header information, we replay them using request generator to get the response header information. We find that between 15% and 40% of Web objects in our traces can not be cached by a Web proxy server. We use a LRU simulator to show the performance gap when the cachability is either considered or not. We show the characteristics of the cachable data set and find that all its characteristics are fairly similar to that of total data set. Finally, we present some additional results for the cachable and total data set: (1) The main reasons for uncachability are: dynamic requests, responses without last-modified header, responses with HTTP "302 Moved Temporarily" status code, and responses with a HTTP/1.0 cookie. (2) The cachability of Web objects can not be ignored in simulation because uncachable objects comprise a huge percentage of the total trace. Simulations without cachability consideration will be misleading.

Fast Learning VIEWNET Architectures for Recognizing 3-D Objects from Multiple 2-D Views

The recognition of 3-D objects from sequences of their 2-D views is modeled by a family of self-organizing neural architectures, called VIEWNET, that use View Information Encoded With NETworks. VIEWNET incorporates a preprocessor that generates a compressed but 2-D invariant representation of an image, a supervised incremental learning system that classifies the preprocessed representations into 2-D view categories whose outputs arc combined into 3-D invariant object categories, and a working memory that makes a 3-D object prediction by accumulating evidence from 3-D object category nodes as multiple 2-D views are experienced. The simplest VIEWNET achieves high recognition scores without the need to explicitly code the temporal order of 2-D views in working memory. Working memories are also discussed that save memory resources by implicitly coding temporal order in terms of the relative activity of 2-D view category nodes, rather than as explicit 2-D view transitions. Variants of the VIEWNET architecture may also be used for scene understanding by using a preprocessor and classifier that can determine both What objects are in a scene and Where they are located. The present VIEWNET preprocessor includes the CORT-X 2 filter, which discounts the illuminant, regularizes and completes figural boundaries, and suppresses image noise. This boundary segmentation is rendered invariant under 2-D translation, rotation, and dilation by use of a log-polar transform. The invariant spectra undergo Gaussian coarse coding to further reduce noise and 3-D foreshortening effects, and to increase generalization. These compressed codes are input into the classifier, a supervised learning system based on the fuzzy ARTMAP algorithm. Fuzzy ARTMAP learns 2-D view categories that are invariant under 2-D image translation, rotation, and dilation as well as 3-D image transformations that do not cause a predictive error. Evidence from sequence of 2-D view categories converges at 3-D object nodes that generate a response invariant under changes of 2-D view. These 3-D object nodes input to a working memory that accumulates evidence over time to improve object recognition. ln the simplest working memory, each occurrence (nonoccurrence) of a 2-D view category increases (decreases) the corresponding node's activity in working memory. The maximally active node is used to predict the 3-D object. Recognition is studied with noisy and clean image using slow and fast learning. Slow learning at the fuzzy ARTMAP map field is adapted to learn the conditional probability of the 3-D object given the selected 2-D view category. VIEWNET is demonstrated on an MIT Lincoln Laboratory database of l28x128 2-D views of aircraft with and without additive noise. A recognition rate of up to 90% is achieved with one 2-D view and of up to 98.5% correct with three 2-D views. The properties of 2-D view and 3-D object category nodes are compared with those of cells in monkey inferotemporal cortex.

Compressive holography of diffuse objects.

We propose an estimation-theoretic approach to the inference of an incoherent 3D scattering density from 2D scattered speckle field measurements. The object density is derived from the covariance of the speckle field. The inference is performed by a constrained optimization technique inspired by compressive sensing theory. Experimental results demonstrate and verify the performance of our estimates.

Restoration of optical objects using regularization

Using the regularization theory for improperly posed problems, we discuss object restoration beyond the diffraction limit in the presence of noise. Only the case of one-dimensional coherent objects is considered. We focus attention n the estimation of the error on the restored objects, and we show that, in most realistic cases, it is at best proportional to an inverse power of

A note on the restoration of optical objects

info:eu-repo/semantics/nonPublished

Supporting students with learning disabilities to explore linear relationships using online learning objects

The study of linear relationships is foundational for mathematics teaching and learning. However, students’ abilities connect different representations of linear relationships have proven to be challenging. In response, a computer-based instructional sequence was designed to support students’ understanding of the connections among representations. In this paper we report on the affordances of this dynamic mode of representation specifically for students with learning disabilities. We outline four results identified by teachers as they implemented the online lessons.

Modelling collisions of rail vehicles with deformable objects

This study utilized the latest computing techniques to analyze the driver's cab of a railroad vehicle colliding with deformable objects. It explored the differences between a collision with a deformable object and a collision with a rigid object. It also examined the differences between a collision with a large simple shaped object and a collision with a life-like object. Tools of analysis included vehicle dynamics analysis and finite element analysis.

Distinguishing between similar tubular objects using pulse reflectometry: a study of trumpet and cornet leadpipes

Acoustic pulse reflectometry is used to reconstruct the internal bore profile of trumpet and cornet leadpipe. The method distinguishes between radii differences as small as 0.03 mm, and has since been used by various UK-based brass instrument manufacturers as a diagnostic tool to detect defects that are significant enough to acoustically alter performance.

Investigating shape-from-shading illusions using solid objects

Recent growth in the shape-from-shading psychophysics literature has been paralled by an increasing availability of computer graphics hardware and software, to the extent that most psychophysical studies in this area now employ computer lighting algorithms. The most widely used of algorithms is shape-from-shading psychophysics is the Phong lighting model. This model, and other shading models of its genre, produce readily ineterpretable imiages of three-dimensional scenes. However, such algorithms are only approximations of how light interacts with real objects in the natural environment. Nevertheless, the results from psychophysical experiments using these techniques have been used to infer the processes underlying the perception of shape-from-shading in natural environments. It is important to establish whether this substitution is ever valid. We report a series of experiments investigating whether two recently reported illusions seen computer-generated, Phond shaded images occur for solid objects under real illuminants. The two illusions investigated are three-dimensional curvature contrast and the illuminant-position effect on perceived curvature. We show that both effects do occur for solid objects, and that the magnitude of these effects are equivalent regardless of whether subjects are presented with ray traced or solid objects.

Molecular Computational Elements Encode Large Populations of Small Objects

Since the introduction of molecular computation1, 2, experimental molecular computational elements have grown3, 4, 5 to encompass small-scale integration6, arithmetic7 and games8, among others. However, the need for a practical application has been pressing. Here we present molecular computational identification (MCID), a demonstration that molecular logic and computation can be applied to a widely relevant issue. Examples of populations that need encoding in the microscopic world are cells in diagnostics or beads in combinatorial chemistry (tags). Taking advantage of the small size9 (about 1 nm) and large 'on/off' output ratios of molecular logic gates and using the great variety of logic types, input chemical combinations, switching thresholds and even gate arrays in addition to colours, we produce unique identifiers for members of populations of small polymer beads (about 100 m) used for synthesis of combinatorial libraries10, 11. Many millions of distinguishable tags become available. This method should be extensible to far smaller objects, with the only requirement being a 'wash and watch' protocol12. Our focus on converting molecular science into technology concerning analog sensors13, 14, turns to digital logic devices in the present work.

Optical colors of 56 near-Earth objects: trends with size and orbit

We present the results of BVRIZ photometry of 56 near-Earth objects (NEOs) obtained with the 1-m Jacobus Kapteyn telescope on La Palma during 2000 and 2001. Our sample includes many NEOs with particularly deep 1 - mum pyroxene/olivine absorption bands, similar to Q-type asteroids. We also classify three NEOs with particularly blue colors. No D-type asteroids were found, placing an upper limit of similar to2% on the fraction of the NEO population originating in the outer main belt or the Trojan clouds. The ratio of dark to bright objects in our sample was found to be 0.40, significantly higher than current theoretical predictions. As well as classifying the NEOs, we have investigated color trends with size and orbit. We see a general trend for larger silicate objects to have shallower absorption bands but find no significant difference in the distribution of taxonomic classes at small and large sizes. Our data clearly show that different taxonomic classes tend to occupy different regions of (a, e) space. By comparing our data with current model predictions for NEO dynamical evolution we see that Q- R-, and V-type NEOs tend to have orbits associated with "fast track" delivery from the main belt, whereas S-type NEOs tend to have orbits associated with "slow track" delivery. This outcome would be expected if space weathering occurs on time scales of >10(6) years. (C) 2003 Elsevier Science (USA). All rights reserved.

Photometry of cometary nuclei: Rotation rates, colours and a comparison with Kuiper Belt Objects

We present time-series data on Jupiter Family Comets (JFCs) 17P/Holmes, 47P/Ashbrook-Jackson and 137P/Shoemaker-Levy 2. In addition we also present results from `snap-shot' observations of comets 43P/Wolf-Harrington, 44P/Reinmuth 2, 103P/Hartley 2 and 104P/Kowal 2 taken during the same run. The comets were at heliocentric distances of between 3 and 7 au at this time. We present measurements of size and activity levels for the snap-shot targets. The time-series data allow us to constrain rotation periods and shapes, and thus bulk densities. We also measure colour indices (V - R) and (R - I) and reliable radii for these comets. We compare all of our findings to date with similar results for other comets and Kuiper Belt Objects (KBOs). We find that the rotational properties of nuclei and KBOs are very similar, that there is evidence for a cut-off in bulk densities at ~0.6 g cm-3 in both populations, and the colours of the two populations show similar correlations. For JFCs, there is no observational evidence for the optical colours being dependent on either position in the orbit or orbital parameters.