High pressure and temperature equation of state and spectroscopic study of CeO2.

One of the most widely used x-ray standards and a highly applied component of catalysis systems, CeO2 has been studied for the purpose of better understanding its equation of state and electronic properties. Diamond anvil cells have been used to extend the equation of state for this material to 130 GPa and explore the electronic behavior with applied load. From the x-ray diffraction studies, it has been determined that the high pressure phase transition extends from approximately 35-75 GPa at ambient temperature. Elevation of temperature is found to decrease the initiation pressure for this transition, with multiple distinct temperature regions which indicate structural related anomalies. In addition, hydrostatic and non-hydrostatic effects are compared and exhibit a drastic difference in bulk moduli. The electronic results indicate a change in the scattering environment of the cerium atom, associated with the high pressure phase transition. Overall, these results present the first megabar pressure study and the first high pressure and temperature study of ceria. Additionally, this shows the first combined study of the K and L III edges of this material to 33 GPa.

Activity modeling using event probability sequences.

Changes in motion properties of trajectories provide useful cues for modeling and recognizing human activities. We associate an event with significant changes that are localized in time and space, and represent activities as a sequence of such events. The localized nature of events allows for detection of subtle changes or anomalies in activities. In this paper, we present a probabilistic approach for representing events using the hidden Markov model (HMM) framework. Using trained HMMs for activities, an event probability sequence is computed for every motion trajectory in the training set. It reflects the probability of an event occurring at every time instant. Though the parameters of the trained HMMs depend on viewing direction, the event probability sequences are robust to changes in viewing direction. We describe sufficient conditions for the existence of view invariance. The usefulness of the proposed event representation is illustrated using activity recognition and anomaly detection. Experiments using the indoor University of Central Florida human action dataset, the Carnegie Mellon University Credo Intelligence, Inc., Motion Capture dataset, and the outdoor Transportation Security Administration airport tarmac surveillance dataset show encouraging results.

The membrane proteins, Spt23p and Mga2p, play distinct roles in the activation of Saccharomyces cerevisiae OLE1 gene expression. Fatty acid-mediated regulation of Mga2p activity is independent of its proteolytic processing into a soluble transcription activator.

The Saccharomyces OLE1 gene encodes the Delta-9 fatty acid desaturase, an enzyme that converts saturated fatty acyl-CoAs into cis-Delta-9 unsaturated fatty acids. OLE1 gene expression is regulated by unsaturated fatty acids, which repress transcription and destabilize the OLE1 mRNA. Expression of OLE1 is activated by N-terminal proteolytic fragments of two homologous endoplasmic reticulum membrane proteins, Spt23p and Mga2p. Disruption of either gene does not significantly affect cell growth or fatty acid metabolism; cells that contain null alleles of both genes, however, are unsaturated fatty acid auxotrophs. An analysis of spt23Delta and mga2Delta strains shows that Spt23p and Mga2p differentially activate and regulate OLE1 transcription. In glucose-grown cells, both genes activate transcription to similar levels of activity. Expressed alone, Mga2p induces high levels of OLE1 transcription in cells exposed to cobalt or grown in glycerol-containing medium. Spt23p expressed alone activates OLE1 transcription to levels similar to those in wild type cells. OLE1 expression is strongly repressed by unsaturated fatty acids in spt23Delta or mga2Delta cells, under all growth conditions. To test if OLE1 expression is controlled by fatty acids at the level of membrane proteolysis, soluble N-terminal fragments of Spt23p and Mga2p that lack their membrane-spanning regions (Deltatm) were expressed under the control of their native promoters in spt23Delta;mga2Delta cells. Under those conditions, Mga2pDeltatm acts as a powerful transcription activator that is strongly repressed by unsaturated fatty acids. By comparison, the Spt23pDeltatm polypeptide weakly activates transcription and shows little regulation by unsaturated fatty acids. Co-expression of the two soluble fragments results in activation to levels observed with the Mga2pDeltatm protein alone. The fatty acid repression of transcription under those conditions is attenuated by Spt23Deltatm, however, suggesting that the two proteins may interact to modulate OLE1 gene expression.

Face verification through tracking facial features.

We propose an algorithm for face verification through tracking facial features by using sequential importance sampling. Specifically, we first formulate tracking as a Bayesian inference problem and propose to use Markov chain Monte Carlo techniques for obtaining an empirical solution. A reparameterization is introduced under parametric motion assumption, which facilitates the empirical estimation and also allows verification to be addressed along with tracking. The facial features to be tracked are defined on a grid with Gabor attributes (jets). The motion of facial feature points is modeled as a global two-dimensional (2-D) affine transformation (accounting for head motion) plus a local deformation (accounting for residual motion that is due to inaccuracies in 2-D affine modeling and other factors such as facial expression). Motion of both types is processed simultaneously by the tracker: The global motion is estimated by importance sampling, and the residual motion is handled by incorporating local deformation into the measurement likelihood in computing the weight of a sample. Experiments with a real database of face image sequences are presented.

Robust structure from motion estimation using inertial data.

The utility of using inertial data for the structure-from-motion (SfM) problem is addressed. We show how inertial data can be used for improved noise resistance, reduction of inherent ambiguities, and handling of mixed-domain sequences. We also show that the number of feature points needed for accurate and robust SfM estimation can be significantly reduced when inertial data are employed. Cramér-Rao lower bounds are computed to quantify the improvements in estimating motion parameters. A robust extended-Kalman-filter-based SfM algorithm using inertial data is then developed to fully exploit the inertial information. This algorithm has been tested by using synthetic and real image sequences, and the results show the efficacy of using inertial data for the SfM problem.

Higher-order-statistics-based detection of vehicles in still images.

We present a statistical pattern recognition scheme for detecting vehicles in still images. The methodology involves pattern classification using higher-order statistics (HOS) in a clustering framework. The proposed method approximately models the unknown distribution of the image patterns of vehicles by learning HOS information about the vehicle class from sample images. Given a test image, statistical information about the background is learned "on the fly." An HOS-based decision measure derived from a series expansion of the multivariate probability density function in terms of the Gaussian function and Hermite polynomials is used to classify test patterns as vehicles or otherwise. Experimental results on real images with cluttered background are given to demonstrate the performance of the proposed method. When tested on real aerial images, the method gives good results, even for complicated scenes. The detection rate is found to be quite good, while the false alarms are very few. The method can serve as an important step toward building an automated traffic monitoring system.

Quantitative analysis of error bounds in the recovery of depth from defocused images.

Depth from defocus involves estimating the relative blur between a pair of defocused images of a scene captured with different lens settings. When a priori information about the scene is available, it is possible to estimate the depth even from a single image. However, experimental studies indicate that the depth estimate improves with multiple observations. We provide a mathematical underpinning to this evidence by deriving and comparing the theoretical bounds for the error in the estimate of blur corresponding to the case of a single image and for a pair of defocused images. A new theorem is proposed that proves that the Cramér-Rao bound on the variance of the error in the estimate of blur decreases with an increase in the number of observations. The difference in the bounds turns out to be a function of the relative blurring between the observations. Hence one can indeed get better estimates of depth from multiple defocused images compared with those using only a single image, provided that these images are differently blurred. Results on synthetic as well as real data are given to further validate the claim.

Noise-resilient estimation of optical flow by use of overlapped basis functions.

Conventional techniques for the computation of optical flow from image gradients are used to formulate the problem as a nonlinear optimization that comprises a gradient constraint term and a field smoothness factor. The results of these techniques are often erroneous, highly sensitive to noise and numerical precision, determined sparsely, and computationally expensive. We regularize the gradient constraint equation by modeling optical flow as a linear combination of a set of overlapped basis functions. We develop a theory for estimating model parameters robustly and reliably. We prove that the extended-least-squares solution proposed here is unbiased and robust to small perturbations in the gradient estimates and to mild deviations from the gradient constraint. Our solution is obtained with a numerically stable sparse matrix inversion, which gives a reliable flow-field estimate over the entire frame. To validate our claims, we perform a series of experiments on standard benchmark data sets at a range of noise levels. Overall, our algorithm outperforms by a wide margin the others considered in the comparison. We demonstrate the applicability of our algorithm to image mosaicking and to motion superresolution through experiments on noisy compressed sequences. We conclude that our flow-field model offers greater accuracy and robustness than conventional optical flow techniques in a variety of situations and permits real-time operation.

Adaptive target detection in foliage-penetrating SAR images using alpha-stable models.

Detecting targets occluded by foliage in foliage-penetrating (FOPEN) ultra-wideband synthetic aperture radar (UWB SAR) images is an important and challenging problem. Given the different nature of target returns in foliage and nonfoliage regions and very low signal-to-clutter ratio in UWB imagery, conventional detection algorithms fail to yield robust target detection results. A new target detection algorithm is proposed that (1) incorporates symmetric alpha-stable (SalphaS) distributions for accurate clutter modeling, (2) constructs a two-dimensional (2-D) site model for deriving local context, and (3) exploits the site model for region-adaptive target detection. Theoretical and empirical evidence is given to support the use of the SalphaS model for image segmentation and constant false alarm rate (CFAR) detection. Results of our algorithm on real FOPEN images collected by the Army Research Laboratory are provided.

Separability-based multiscale basis selection and feature extraction for signal and image classification.

Algorithms for multiscale basis selection and feature extraction for pattern classification problems are presented. The basis selection algorithm is based on class separability measures rather than energy or entropy. At each level the "accumulated" tree-structured class separabilities obtained from the tree which includes a parent node and the one which includes its children are compared. The decomposition of the node (or subband) is performed (creating the children), if it provides larger combined separability. The suggested feature extraction algorithm focuses on dimensionality reduction of a multiscale feature space subject to maximum preservation of information useful for classification. At each level of decomposition, an optimal linear transform that preserves class separabilities and results in a reduced dimensional feature space is obtained. Classification and feature extraction is then performed at each scale and resulting "soft decisions" obtained for each area are integrated across scales. The suggested algorithms have been tested for classification and segmentation of one-dimensional (1-D) radar signals and two-dimensional (2-D) texture and document images. The same idea can be used for other tree structured local basis, e.g., local trigonometric basis functions, and even for nonorthogonal, redundant and composite basis dictionaries.

Region adaptive subband image coding.

We present a region adaptive subband image coding scheme using the statistical properties of image subbands for various subband decompositions. Motivated by analytical results obtained when the input signal to the subband decomposition is a unit step function, we analyze the energy packing properties toward the lower frequency subbands, edges, and the dependency of energy distribution on the orientation of the edges, in subband decomposed images. Based on these investigations and ideal analysis/synthesis filtering done in the frequency domain, the region adaptive subband image coding scheme extracts suitably shaped regions in each subband and then uses adaptive entropy-constrained quantizers for different regions under the assumption of a generalized Gaussian distribution for the image subbands. We also address the problem of determining an optimal subband decomposition among all possible decompositions. Experimental results show that visual degradations in the reconstructed image are negligible at a bit rate of 1.0 b/pel and reasonable quality images are obtainable at rates as low as 0.25 b/pel.

Probe-based automatic target recognition in infrared imagery.

A probe-based approach combined with image modeling is used to recognize targets in spatially resolved, single frame, forward looking infrared (FLIR) imagery. A probe is a simple mathematical function that operates locally on pixel values and produces an output that is directly usable by an algorithm. An empirical probability density function of the probe values is obtained from a local region of the image and used to estimate the probability that a target of known shape is present. Target shape information is obtained from three-dimensional (3-D) computer-aided design (CAD) models. Knowledge of the probe values along with probe probability density functions and target shape information allows the likelihood ratio between a target hypothesis and background hypothesis to be written. The generalized likelihood ratio test is then used to accept one of the target poses or to choose the background hypothesis. We present an image model for infrared images, the resulting recognition algorithm, and experimental results on three sets of real and synthetic FLIR imagery.

Multiresolution Gauss-Markov random field models for texture segmentation.

This paper presents multiresolution models for Gauss-Markov random fields (GMRFs) with applications to texture segmentation. Coarser resolution sample fields are obtained by subsampling the sample field at fine resolution. Although the Markov property is lost under such resolution transformation, coarse resolution non-Markov random fields can be effectively approximated by Markov fields. We present two techniques to estimate the GMRF parameters at coarser resolutions from the fine resolution parameters, one by minimizing the Kullback-Leibler distance and another based on local conditional distribution invariance. We also allude to the fact that different GMRF parameters at the fine resolution can result in the same probability measure after subsampling and present the results for first- and second-order cases. We apply this multiresolution model to texture segmentation. Different texture regions in an image are modeled by GMRFs and the associated parameters are assumed to be known. Parameters at lower resolutions are estimated from the fine resolution parameters. The coarsest resolution data is first segmented and the segmentation results are propagated upward to the finer resolution. We use the iterated conditional mode (ICM) minimization at all resolutions. Our experiments with synthetic, Brodatz texture, and real satellite images show that the multiresolution technique results in a better segmentation and requires lesser computation than the single resolution algorithm.

A spectral attentional mechanism tuned to object configurations.

This paper describes an attentional mechanism based on the interpretation of spectral signatures for detecting regular object configurations in areas of an image delineated using context information. The proposed global operator relies on the spectral analysis of edge structure and exploits spatial as well as frequency-domain constraints derived from known geometrical models of monitored objects. A decision theoretic method for learning acceptance detection regions is presented. Applications of this attentional mechanism are demonstrated for several aerial image interpretation tasks for attentional as well as recognition purposes. Specific examples are described for detecting vehicle formations (such as convoys), qualifying the geometry of detected formations, or monitoring the occupancy of regions of interest (such as parking areas, roads, or open areas). Experiments and sensitivity analysis results are reported.

Simulation of error in optical radar range measurements.

We describe a computer simulation of atmospheric and target effects on the accuracy of range measurements using pulsed laser radars with p-i-n or avalanche photodiodes for direct detection. The computer simulation produces simulated images as a function of a wide variety of atmospheric, target, and sensor parameters for laser radars with range accuracies smaller than the pulse width. The simulation allows arbitrary target geometries and simulates speckle, turbulence, and near-field and far-field effects. We compare simulation results with actual range error data collected in field tests.

Delineating buildings by grouping lines with MRFs.

Traditionally, Markov random field (MRF) models have been used in low-level image analysis. The article presents an MRF-based scheme to perform object delineation. The proposed edge-based approach involves extracting straight lines from the edge map of an image. Then, an MRF model is used to group these lines to delineate buildings in aerial images.

Tracking a dynamic set of feature points.

We address the problems of tracking a set of feature points over a long sequence of monocular images as well as how to include and track new feature points detected in successive frames. Due to the 3-D movement of the camera, different parts of the images exhibit different image motion. Tracking discrete features can therefore be decomposed into several independent and local problems. Accordingly, we propose a localized feature tracking algorithm. The trajectory of each feature point is described by a 2-D kinematic model. Then to track a feature point, an interframe motion estimation scheme is designed to obtain the estimates of interframe motion parameters. Subsequently, using the estimates of motion parameters, corresponding points are identified to subpixel accuracy. Afterwards, the temporal information is processed to facilitate the tracking scheme. Since different feature points are tracked independently, the algorithm is able to handle the image motion arising from general 3-D camera movements. On the other hand, in addition to tracking feature points detected at the beginning, an efficient way to dynamically include new points extracted in subsequent frames is devised so that the information in a sequence is preserved. Experimental results for several image sequences are also reported.

Scalable data parallel algorithms for texture synthesis using Gibbs random fields.

This article introduces scalable data parallel algorithms for image processing. Focusing on Gibbs and Markov random field model representation for textures, we present parallel algorithms for texture synthesis, compression, and maximum likelihood parameter estimation, currently implemented on Thinking Machines CM-2 and CM-5. The use of fine-grained, data parallel processing techniques yields real-time algorithms for texture synthesis and compression that are substantially faster than the previously known sequential implementations. Although current implementations are on Connection Machines, the methodology presented enables machine-independent scalable algorithms for a number of problems in image processing and analysis.

A unified approach to boundary perception: edges, textures, and illusory contours.

A model consisting of a multistage system which extracts and groups salient features in the image at different spatial scales (or frequencies) is used. In the first stage, a Gabor wavelet decomposition provides a representation of the image which is orientation selective and has optimal localization properties in space and frequency. This decomposition is useful in detecting significant features such as step and line edges at different scales and orientations in the image. Following the wavelet transformation, local competitive interactions are introduced to reduce the effects of noise and changes in illumination. Interscale interactions help in localizing the line ends and corners, and play a crucial role in boundary perception. The final stage groups similar features, aiding in boundary completion. The different stages can be identified with processing by simple, complex, and hypercomplex cells in the visual cortex of mammals. Experimental results demonstrate the performance of this model in detecting boundaries (both real and illusory) in real and synthetic images.

VLSI neuroprocessors for video motion detection.

The system design of a locally connected competitive neural network for video motion detection is presented. The motion information from a sequence of image data can be determined through a two-dimensional multiprocessor array in which each processing element consists of an analog neuroprocessor. Massively parallel neurocomputing is done by compact and efficient neuroprocessors. Local data transfer between the neuroprocessors is performed by using an analog point-to-point interconnection scheme. To maintain strong signal strength over the whole system, global data communication between the host computer and neuroprocessors is carried out in a digital common bus. A mixed-signal very large scale integration (VLSI) neural chip that includes multiple neuroprocessors for fast video motion detection has been developed. Measured results of the programmable synapse, and winner-takes-all circuitry are presented. Based on the measurement data, system-level analysis on a sequence of real-world images was conducted.