Search

This paper presents a robust method for number plate detection, where mean shift segmentation is used to segment color vehicle images into candidate regions. Three features are extracted in order to decide whether a candidate region contains a number plate, namely, rectangularity, aspect ratio, and edge density. Then, the Mahalanobis classifier is used with respect to the above three features to detect number plate regions accurately. The experimental results show that our algorithm produces high robustness and accuracy. © 2005 IEEE.

The one-sided two-level crossing reduction problem is an important problem in hierarchical graph drawing. Because of its NP-hardness there are many heuristics, such as the well-known barycenter and median heuristics. We consider the constrained one-sided two-level crossing reduction problem, where the relative position of certain vertex pairs on the second level is fixed. Based on the barycenter heuristic, we present a new algorithm that runs in quadratic time and generates fewer crossings than existing simple extensions. It is significantly faster than an advanced algorithm by Schreiber [12] and Finnocchi [1,2,6], while it compares well in terms of crossing number. It is also easy to implement.

10 pages; International audience; Animal distribution and abundance are greatly affected by the availability of their food resources, which also depends on landscape structure. Lothar hurricane in 1999 had profoundly modified the structure of the forests in France, affecting the habitat quality of ungulates. We tested whether the variations in home-range size of 23 female roe deer were influenced by the fragmentation of the landscape caused by Lothar in the Chize´ forest, namely by the increase in heterogeneity associated with the localized massive tree felling. Home-range size was studied in the summers of 2001 and 2002 and we found that variation in home-range size was mainly explained by only one landscape variable: edge density. Home-range size decreased as edge density increased, which is consistent with the fact that edges are good browsing habitats for roe deer. The result of this study suggests that, after 2 years, the hurricane had improved the quality of the home ranges by creating more forest heterogeneity and increasing the contacts between the different vegetation patches within the home range. These results highlight the fact that spatial heterogeneity is likely to be a key factor influencing the distribution and local population density.

We grew SrTiO3 on SrTiO3 [001] by pulsed laser deposition, while observing x-ray diffraction at the (0 0 .5) position. The drop dI in the x-ray intensity following a laser pulse contains information about plume-surface interactions. Kinematic theory predicts dI/I = -4sigma(1-sigma), so that dI/I depends only on the amount of deposited material sigma. In contrast, we observed experimentally that |dI/I| < 4sigma(1-sigma), and that dI/I depends on the phase of x-ray growth oscillations. The combined results suggest a fast smoothing mechanism that depends on surface step-edge density., 4 figures

Spatio-temporal variations in conspecific density and resource availability are two of the main factors responsible for plasticity in habitat selection. Despite the need for habitat selection models that can accurately predict animal distribution given the plasticity in the selection process, no study has assessed the synergistic effects of these factors on habitat selection. We investigated density-dependent functional responses by raccoons (Procyon lotor) and striped skunks (Mephitis mephitis), two of the main hosts of the rabies virus in North America. We monitored 54 raccoons and 12 striped skunks with Global-Positioning-System collars in a landscape dominated by corn fields and forest patches. We built resource selection functions to evaluate if the selection of corn fields varied with conspecific density and corn field availability within 100% minimum convex polygons. Raccoons altered their selection of corn fields depending on both conspecific density and corn-forest edge density or corn field prop...

We evaluated various texture features and region of interest (ROI) types of breast ultrasonograms in order to determine the best-performing combinations for differentiating between benign and malignant solid breast nodules. A total of 21 breast ultrasonograms (12 benign, nine malignant) containing solid breast nodules were evaluated. Eight ROI types were defined around the nodules. The texture feature of each ROJ was measured and the ratios of texture features were calculated for each pair of ROIs. This procedure was repeated for five different feature types, thus yielding texture feature ratios for 140 different combinations of ROIs and texture features. We evaluated the performance of the texture feature ratio in differentiating between benign and malignant nodules using t test analysis. Evaluating the top ranked texture and ROI combinations, we found edge density and mutual information were the best two texture features, and that the ROI types of outside lesion and lesion margin had good performance.

Abstract We propose a simple algorithm based on a Markov chain for generating acyclic digraphs with a given number of vertices uniformly at random. In addition, standard combinatorial techniques allow us to describe the overall shape and average edge density of an acyclic digraph. As an application, we were able to estimate the density function of classical statistics on acyclic digraphs.

The HHFW system on NSTX has operated with the full 12-antenna, 6-transmitter configuration, delivering over 2 MW reliably for pulse lengths over 100 ms with various phasings of the antenna system. A circuit model of the full 12- antenna coupled system has been developed that gives good agreement with vacuum measurements. When it is used to extract the effects of the plasma on the rf circuit, pronounced asymmetries in antenna loading are observed, even when antenna phasing is symmetrical (e.g., 0π0π0π…..). The loading of the plasma on the antenna has been calculated with the RANT3D code using measured edge density profiles in front of the antenna; these agree with measured loading values.

Although many people have tried using genetic algorithms (GAs) for exam timetabling, far fewer have done systematic investigations to try to determine whether a GA is a good choice of method or not. We have extensively studied GAs that use one particular kind of direct encoding for exam timetabling. Perhaps not surprisingly, it emerges that this approach is not very good, but it is instructive to see why. In the course of this investigation we discovered a class of solvable problems with interesting properties: our GAs would sometimes fail to solve some of the moderately-constrained problems, but could solve all of the lightly-constrained ones and all of the highly-constrained ones. This is despite the fact that they form a hierarchy: those erratically-solved problems are subproblems of the easily-solved but highly-constrained ones. Moreover, some other non-evolutionary approaches also failed on precisely the same sets. This, together with some observations about much simpler graph-colouring methods based on the Brelaz algorithm, suggest some future directions for GA-based methods.

A major emphasis in landscape ecology is the study of landscape pattern, including pattern dynamics, ecological processes that influence pattern, and effects of pattern on organisms. In order to investigate relationships between landscape pattern and ecological processes, it is often helpful to describe the patterns in quantifiable terms, and a variety of spatial metrics have been developed for this purpose (Whitcomb et al. 1981 Forman and Godron 1986 O’Neill et al. 1988 Turner 1990 Milne 1991 Gustafson and Parker 1992] Li and Reynolds 1993 Plotnick et al. 1993 McGarigal and Marks 1995). Although these metrics enable ecologists to address landscape-level questions in a more rigorous fashion, they often are applied without a clear understanding of the strengths and limitations of each measure or how various measures are interrelated.

Ion Bernstein wave (IBW) antennas are known to have substantial localized effects on the plasma edge. To allow better understanding and measurement of these effects, the TFTR edge reflectometer has been relocated to the new IBW antenna. This move was facilitated by the incorporation of a diagnostic access tube in the IBW antenna identical to the original diagnostic tube in the fast-wave (FW) antenna. This allowed the reflectometer launcher to simply be moved from the old FW antenna to the new IBW antenna. Only a moderate extension of the waveguide transmission line was required to reconnect the reflectometer to the launcher in its new location. Edge density profile modification during IBW experiments has been observed. Results from IBW experiments will be presented and contrasted to the edge density modifications previously observed during FW heating experiments.

The build-up of plasma parameters following the H-mode transition in JET has been analysed in view of its consequences for the alpha power evolution in the access to burning plasma conditions in ITER. JET experiments show that the build-up of plasma temperature both at the plasma core and the plasma edge occurs in timescales comparable to the energy confinement time. In contrast, the evolution of the edge and core densities differs strongly depending on the level of plasma current in the discharge and of the associated NBI penetration. For higher plasma current H-mode discharges (Ip > 2.0–2.5 MA, depending on plasma shape), with naturally higher plasma densities for which NBI penetration is poorer, the core density evolves in much longer timescales than the edge density leading to the formation of rather hollow density profiles. These hollow density profiles persist for timescales of several energy confinement times until they are usually terminated by a sawtooth. Modelling of the JET experiments with JETTO shows that the density build-up following the H-mode transition can be described with a purely diffusive model, despite the low collisionalities of high current H-mode plasmas at JET. The consequences of these JET experimental/modelling findings for the access to burning plasma conditions in the ITER QDT = 10 scenario are presented.

Features of L-H transition are investigated using the supersonic molecular beam injection (SMBI) with directional velocity under low heating powers in the present paper. Comparing the edge density profiles, it is analyzed that the conventional gas puffing (GP) and SMBI have effects on L-H transition. Experimental results suggest that the SMBI does directly trigger L-H transition on HL-2A, and it considerably reduces the power threshold of L-H transition. After abundant experimental data are analyzed and processed, it is found that the minimum L-H transition power with SMBI decreases by 10% when it is compared with GP trigged H-mode under same conditions.

We present and analyze a portable, high-performance algorithm for finding connected components on modern distributed memory multiprocessors. The algorithm is a hybrid of the classic DFS on the subgraph local to each processor and a variant of the Shiloach-Vishkin PRAM algorithm on the global collection of subgraphs. We implement the algorithm in Split-C and measure performance on the the Cray T3D, the Meiko CS-2, and the Thinking Machines CM-5 using a class of graphs derived from cluster dynamics methods in computational physics. On a 256 processor Cray T3D, the implementation outperforms all previous solutions by an order of magnitude. A characterization of graph parameters allows us to select graphs that highlight key performance features. We study the effects of these parameters and machine characteristics on the balance of time between the local and global phases of the algorithm and find that edge density, surface-to-volume ratio, and relative communication cost dominate performance. By understanding the effect of machine characteristics on performance, the study sheds light on the impact of improvements in computational and/or communication performance on this challenging problem.

Transport of metallic impurities in the plasma edge was compared for normal and detached Tokamak plasmas. A neo-classical model, valid for a dirty plasma in the Pfirsch-Schluter regime, was used to calculate the impurity flux terms. Changes in the edge density and temperature profiles, characteristic of detached discharges, were seen to greatly influence the impurity transport. In particular, the inward convective velocity increased significantly because of the detachment. A one-dimensional transport code was used to evaluate the influence of the edge transport modifications on the global impurity confinement time. Code predictions are in qualitative agreement with experimental measurements made on TEXTOR.

It is shown that for arbitrary positivee there exists a graph withoutK 4 and so that all its subgraphs containing more than 1/2 +e portion of its edges contain a triangle (Theorem 2). This solves a problem of Erdos and Nesetřil. On the other hand it is proved that such graphs have necessarily low edge density (Theorem 4). Theorem 3 which is needed for the proof of Theorem 2 is an analog of Goodman's theorem [8], it shows that random graphs behave in some respect as sparse complete graphs. Theorem 5 shows the existence of a graph on less than 1012 vertices, withoutK 4 and which is edge-Ramsey for triangles.

This paper deals with digital patterns generated by a class of random geometric processes proposed as models for images. The joint gray-level probability density for pairs of points at given separations is used to derive various second-order image statistics, including the autocorrelation, edge density, and variogram. In modeling images using these processes, these results can be used for fitting models to a given ensemble of images.

For the divertor configurations of ASDEX a two-fluid model describing the plasma flow and energy transport in the scrape-off layer together with models for the neutral particle dynamics in the divertor chamber have been used to study the relation between the neutral molecule density and plasma density in the divertor and the scrape-off density at the midplane. Results of DEGAS Monte Carlo calculations prove the divertor entrance slits to be nearly impenetrable for neutrals in the presence of a plasma. A global recycling picture is suggested relying on edge density control by the divertor and a hydrogen ion inward drift in the main discharge region to maintain the bulk plasma density even in the absence of neutrals in the main chamber.

Let G be a finite simple graph on n vertices with minimum degreeδ(G) ⩾ δ (n ≡ δ (mod 2)). Let max (k, G) denote the set of all k-subsets A ⊆ V (G) such that the number of edges in the induced subgraph 〈A〉 is a maximum. We prove that for some i ϵ{ 0, 1.2, …, ⌊ 1 2 δ⌋} there exist a partition (X, Y) of V(G) such that (i) (X) = ⌜ 1 2 n⌝ + i, |Y| = ⌊ 1 2 n⌋ − i ; (ii) δ(X)⩾ ⌜ 1 2 δ⌝ + i δ(Y) ⩾ ⌊ 1 2 δ⌋ −i ; (iii) Xϵ max(⌜ 1 2 n⌝ + i, G) or Y ϵ max (⌊ 1 2 n⌋ − i, G) . Analogous edge density constraints, rather than constraints on the minimum degree of G, guaranteeing such a partition are also discussed.

Dispersion relations for Alfven surface waves in a cylindrical plasma have previously been obtained by assuming that the plasma is separated from a conducting wall by a vacuum gap. The authors examine the effect of replacing the vacuum layer by a region of low density plasma. The low density region has little effect on negative m surface waves, but introduces an ion cyclotron resonance and a waveguide cut-off for positive m modes. These effects are associated with the fact that positive m modes are left hand polarised near the plasma edge. Dispersion relations for stepped and parabolic density profiles are examined. A profile with N discrete steps and a vacuum boundary supports a discrete spectrum of N+1 first radial modes. When N is large, these modes can be identified as N localised torsional wave modes plus one surface wave mode.

This paper gives new constructions of k -chromatic critical graphs with high minimum degree and high edge density, and of vertex-critical graphs with high edge density.

A simple polynomial-time algorithm is presented which computes independent sets of guaranteed size in connected triangle-free noncubic graphs with maximum degree 3. Let n and m denote the number of vertices and edges, respectively, and let c = m/n denote the edge density where c < 3/2. The algorithm establishes new lower bounds on the independence ratio of these graphs for 1 < c < 3/2.

Improving a recent result of G. Koester, it is shown that there exist 4-critical planar graphs in which the number of edges is more than twice the number of vertices.

Visually evoked potentials were recorded from bipolar electrodes implanted in the foveal and parafoveal striate, prestriate and inferotemporal cortices of mature rhesus monkeys (Macaca mulatta). The stimuli used were a blank field and three checkerboard patterns containing checks whose sizes were 2° 24′, 36′, and 9′ of visual arc. In addition to the size of the checks within the patterns, the intensity at which the stimuli were presented was also varied. In all four regions studied, check size had a significant effect on the amplitude of the average evoked potential, but intensity showed an effect only in the two striate regions. Certain regions of the visual cortex can apparently process two dimensions of a stimulus concurrently, and the same mass of neuronal tissue can participate in the processing of different dimensions of a stimulus at different times. The results are discussed in the context of existing literature concerning the evoked potential as an indicator of sensory information processing.

Superconcentration is a strong property of interconnection diagraphs. We characterize its negation by existence of two disjoint and seperated sets which shrink under the forward and backward neighbor relation, respectively. This enables a better, non-cascaded design of superconcentrators, explicit ones with edge density ≤ 118, random ones with edge density ≤ 13.

JET ITER-like wall (ILW) experiments show that the edge density evolution is strongly linked with the poloidal distribution of the ionization source. The fueling profile in the JET-ILW is more delo ...

By varying growth temperature, ab abrupt change in density and absorption edge energy is produced in amorphous Ge films. For 25 – 200°C the absorption edge occurs at 0.6 eV and the density is less than crystalline Ge. For 250 – 300°C the edge occurs near 0.8 eV and the edge density is close to the crystalline density.