Dimitrios Charalampidis

Wavelet-based rotational invariant roughness features for texture classification and segmentation

In this paper, we introduce a rotational invariant feature set for texture segmentation and classification, based on an extension of fractal dimension (FD) features. The FD extracts roughness information from images considering all available scales at once. In this work, a single scale is considered at a time so that textures with scale-dependent properties are satisfactorily characterized. Single-scale features are combined with multiple-scale features for a more complete textural representation. Wavelets are employed for the computation of single- and multiple-scale roughness features because of their ability to extract information at different resolutions. Features are extracted in multiple directions using directional wavelets, and the feature vector is finally transformed to a rotational invariant feature vector that retains the texture directional information. An iterative K-means scheme is used for segmentation, and a simplified form of a Bayesian classifier is used for classification. The use of the roughness feature set results in high-quality segmentation performance. Furthermore, it is shown that the roughness feature set exhibits a higher classification rate than other feature vectors presented in this work. The feature set retains the important properties of FD-based features, namely insensitivity to absolute illumination and contrast.

Oxidative stress contributes to methamphetamine-induced left ventricular dysfunction

AIMS Our aim was to test the hypothesis that the repeated, binge administration of methamphetamine would produce oxidative stress in the myocardium leading to structural remodeling and impaired left ventricular function. METHODS AND RESULTS Echocardiography and Millar pressure-volume catheters were used to monitor left ventricular structure and function in rats subjected to four methamphetamine binges (3 mg/kg, iv for 4 days, separated by a 10-day drug-free period). Hearts from treated and control rats were used for histological or proteomic analysis. When compared with saline treatment, four methamphetamine binges produced eccentric left ventricular hypertrophy. The drug also significantly impaired systolic function (decreased fractional shortening, ejection fraction, and adjusted maximal power) and produced significant diastolic dysfunction (increased -dP/dt and tau). Dihydroethedium staining showed that methamphetamine significantly increased (285%) the levels of reactive oxygen species in the left ventricle. Treatment with methamphetamine also resulted in the tyrosine nitration of myofilament (desmin, myosin light chain) and mitochondrial (ATP synthase, NADH dehydrogenase, cytochrome c oxidase, prohibitin) proteins. Treatment with the superoxide dismutase mimetic, tempol in the drinking water prevented methamphetamine-induced left ventricular dilation and systolic dysfunction; however, tempol (2.5 mM) did not prevent the diastolic dysfunction. Tempol significantly reduced, but did not eliminate dihydroethedium staining in the left ventricle, nor did it prevent the tyrosine nitration of mitochondrial and contractile proteins. CONCLUSION This study shows that oxidative stress plays a significant role in mediating methamphetamine-induced eccentric left ventricular dilation and systolic dysfunction.

Texture synthesis: textons revisited

This paper introduces a technique for synthesizing natural textures, with emphasis on quasiperiodic and structural textures. Textures are assumed to be composed of three components, namely illumination, structure, and stochastic. The contribution of this work is that, in contrast to previous techniques, it proposes a joint approach for handling the textures global illumination, irregular structure, and stochastic component which may be correlated to the other two components. Furthermore, the proposed technique does not produce verbatim copies in the synthesized texture. More specifically, a top-down approach is used for extraction of texture elements (textons) in which, in contrast to previous texton-based approaches, no assumptions regarding perfect periodicity are made. The structure itself can be modeled as a stochastic process. Consequently, textons are allowed to have irregular and nonidentical shapes. In the synthesis stage, a new nonregular textural structure is designed from the original one that defines the place holders for textons. We call such place holders empty textons (e-textons). The e-textons are filled in by a representative texton. Since e-textons do not have identical shapes, a texton shape-matching procedure is required. After adding the illumination to the structural component, a strictly localized version of a block sampling technique is applied to add the stochastic component. The block sampling technique combined with the addition of the illumination component provides a significant improvement in the appearance of synthesized textures. Results show that the proposed method is successful in synthesizing structural textures visually indistinguishable to the original. Moreover, the method is successful in synthesizing a variety of stochastic textures.

Steerable Weighted Median Filters

A filter is steerable if transformed (i.e., rotated, scaled, etc.) versions of its impulse response can be expressed as linear combinations of a fixed set of basis functions. Steerability is important for numerous image processing applications. However, it is a property presently shared only by a specific class of linear filters. On the other hand, several classes of nonlinear filters, such as weighted median filters (WMFs), may offer certain advantages over linear filters such as robustness and edge preserving capabilities. In this paper, the concept of steerability is extended to encompass WMFs. It will be shown that, in general, a steerable WMF design technique needs to be capable of handling negative weights. Although methods that allow the design of WMFs admitting negative weights have already been proposed, such methods do not necessarily produce filters that are steerable, as opposed to the approach presented in this work. Experimental results illustrate the applicability of steerable WMFs in two applications, namely edge detection and orientation analysis.

Efficient Directional Gaussian Smoothers

Linear and nonlinear filters, including morphological operators, play a significant role in the processing of remote sensing imagery. In particular, smoothing filters have been extensively used for noise removal and image restoration. In applications where linear and shift-invariant filters can be effectively employed, filtering is computationally efficient if implemented in transform domains. Nevertheless, in remote sensing applications, it is essential that smoothing filters be capable of handling missing and erroneous data without loss of information. In such cases, filtering requires the involvement of logical operations in order to determine which pixels should be used for processing, and thus takes the form of a nonlinear operator. Hence, transform-based methods cannot be used. Still, in applications where large volumes of data need to be processed, it is greatly desired that fast filtering algorithms are used. This letter introduces a computationally efficient spatial-domain-based implementation which is partially separable and steerable. The technique is general, and its efficiency has been demonstrated on weather radar data. It is shown that the proposed filtering approach is significantly faster compared to a recently introduced separable filter implementation.

Efficient FPGA implementation of steerable Gaussian smoothers

Smoothing filters have been extensively used in image and video analysis. In particular, directional smoothers have been employed in motion analysis, edge detection, line parameter estimation, and texture analysis. Such applications often necessitate the use of several directional filters oriented at different angles. However, applying a large number of filters commonly requires a significant amount of computing resources. In such cases, real-time performance may be possibly achieved through utilization of hardware devices having parallel processing capabilities. Additionally, techniques can take advantage of the inherent properties of certain smoothing filters. Such a property is steerability, which implies that the outputs of several filtering operations can be linearly combined in order to produce the output of a directional filter at an arbitrary orientation. Although several efficient FPGA implementations of the convolution operation have been presented in the literature for non-separable and separable, research on steerable filter implementations on FPGA is limited. In this paper, steerable Gaussian smoothers are implemented on an FPGA platform. The technique is compared with a software-based implementation. Performance comparisons indicate that the FPGA technique provides significant speed-up factor of at least ~6, utilizing only a small percentage of the FPGA resources.

Sensor Management for Collision Alert in Orbital Object Tracking

Given the increasingly dense environment in both low-earth orbit (LEO) and geostationary orbit (GEO), a sudden change in the trajectory of any existing resident space object (RSO) may cause potential collision damage to space assets. With a constellation of electro-optical/infrared (EO/IR) sensor platforms and ground radar surveillance systems, it is important to design optimal estimation algorithms for updating nonlinear object states and allocating sensing resources to effectively avoid collisions among many RSOs. Previous work on RSO collision avoidance often assumes that the maneuver onset time or maneuver motion of the space object is random and the sensor management approach is designed to achieve efficient average coverage of the RSOs. Few attempts have included the inference of an objects intent in the response to an RSOs orbital change. We propose a game theoretic model for sensor selection and assume the worst case intentional collision of an objects orbital change. The intentional collision results from maximal exposure of an RSOs path. The resulting sensor management scheme achieves robust and realistic collision assessment, alerts the impending collisions, and identifies early RSO orbital change with lethal maneuvers. We also consider information sharing among distributed sensors for collision alert and an objects intent identification when an orbital change has been declared. We compare our scheme with the conventional (non-game based) sensor management (SM) scheme using a LEO-to-LEO space surveillance scenario where both the observers and the unannounced and unplanned objects have complete information on the constellation of vulnerable assets. We demonstrate that, with adequate information sharing, the distributed SM method can achieve the performance close to that of centralized SM in identifying unannounced objects and making early warnings to the RSO for potential collision to ensure a proper selection of collision avoidance action.

FPGA implementation of graph cut based image thresholding

Thresholding is an important process in many image processing applications. Recently, a bi-level image thresholding method based on graph cut was proposed. The method provided thresholding results which were superior to those obtained with previous techniques. Moreover, the technique was computationally less complex compared to other graph cut-based image thresholding approaches. However, the execution time requirements may still be significant, especially if it is of interest to perform real-time thresholding of a large number of images, such as in the case of high-resolution video sequences. In this paper, we propose a method based on the previously proposed graph cut thresholding method, which is nevertheless appropriate for hardware (FPGA) real-time implementations. A subset of the proposed modifications are also appropriate for a general software implementation. Considering only this subset, the C implementation of the modified method is approximately 2.2 times faster than the original method, as it was presented in the original graph cut-based thresholding paper. Furthermore, the FPGA-based implementation is designed to be 70-100 times faster than the software implementation, depending on the image used.

Median filter on FPGAs

Median filters have been shown to be effective in removing impulsive noise from images. This paper proposes three different directional median filter implementations on FPGA. The techniques are capable of performing median filtering operations for four different directions simultaneously. These implementations are extensions to an existing cumulative histogram-based median filtering technique. Yet, directional processing was not addressed in previous work. Directional median filters process images along the direction of the filtering window, and have been shown to be more effective in removing impulsive noise compared to traditional median filters. All techniques aimed to decrease processing time, while reducing the resource utilization on hardware. Experimental results confirmed that all three techniques are synthesizable and implementable on Xilinx Virtex 5. There is a trade-off between resource utilization and processing time among the three techniques.

Tracking of storm fronts in weather radar imagery

Tracking of storm fronts in weather imagery is important for several weather-related applications. Coastal-area weather radars provide coverage up to 200-250 miles into the ocean, and thus can help with tracking of storm-fronts to support forecasting in those areas. Another application where tracking of storm fronts can be of assistance is clutter/rain classification. Specifically, the path of a tracked event can be used to decide if the particular event corresponds to precipitation or clutter. For instance, clutter usually appears to be a relatively static event. Precipitation can be modeled as a mixture of localized functions, each changing in terms of shape, position, and intensity. Tracking of precipitation events can be performed via tracking of the localized function parameters. In this paper, the modeling of rain events using Radial Basis Function neural networks (RBFNN) is studied. In the recent past, such techniques have been used for forecasting. Although effective, these techniques have been found to be computationally expensive. In this work, we evaluate the feasibility of modeling rain events using RBFNN in an efficient manner, and we propose modifications to existing techniques to achieve this goal.