108,381 research outputs found
Edge detection filter based on Mumford-Shah green function
No full textIn this paper, we propose an edge detection algorithm based on the Green function associated with Mumford-Shah (M-S) segmentation model. This Green function has a singularity at its center. A regularization method is therefore proposed here to obtain an edge detection filter known here as Bessel filter. This filter is robust in the presence of noise and its implementation is simple. It is demonstrated here that this filter detects edges particularly in the case of curved boundaries and sharp corners, more accurately than popular filters in the recent literature. A mathematical argument is also provided to prove that the gradient magnitude of the convolved image with this filter has local maxima in discontinuities of the original image. The Bessel filter enjoys better overall performance (the product of the detection performance and localization indices) in Canny-like criteria than the state of art filters in the literature. Quantitative and qualitative evaluations of the edge detection algorithms investigated in this paper on synthetic and real world benchmark images confirm the theoretical results presented here, indicating the superiority of the Bessel filter over the popular edge detection filters. The numerical complexity of the algorithm proposed here is as low as any convolution-based edge detection algorithm
Airfoil noise reductions through leading edge serrations
No full textThis paper provides an experimental investigation into the use of leading edge (LE) serrations as a means of reducing the broadband noise generated due to the interaction between the aerofoil’s LE and impinging turbulence. Experiments are performed on a flat plate in an open jet wind tunnel. Grids are used to generate isotropic homogeneous turbulence. The leading edge serrations are in the form of sinusoidal profiles of wavelengths, ?, and amplitudes, 2h. The frequency and amplitude characteristics are studied in detail in order to understand the effect of LE serrations on noise reduction characteristics and are compared with straight edge baseline flat plates. Noise reductions are found to be insignificant at low frequencies but significant in the mid frequency range (500 Hz–8 kHz) for all the cases studied. The flat plate results are also compared to the noise reductions obtained on a serrated NACA-65 aerofoil with the same serration profile. Noise reductions are found to be significantly higher for the flat plates with a maximum noise reduction of around 9 dB compared with about 7 dB for the aerofoil. In general, it is observed that the sound power reduction level (?PWL) is sensitive to the amplitude, 2h of the LE serrations but less sensitive to the serration wavelength, ?. Thus, this paper sufficiently demonstrates that the LE amplitude acts as a key parameter for enhancing the noise reduction levels in flat plates and aerofoils
Direct numerical simulation of turbulent flow past a trailing edge and the associated noise generation
No full textDirect numerical simulations (DNS) are conducted of turbulent flow passing an infinitely thin trailing edge (TE). The objective is to investigate the turbulent flow field in the vicinity of the TE and the associated broadband noise generation. To generate a turbulent boundary layer a short distance from the inflow boundary, high amplitude lifted streaks and disturbances that can be associated with coherent outer layer vortices are introduced at the inflow boundary. A rapid increase in skin friction and a decrease in boundary layer thickness and pressure fluctuations is observed at the trailing edge. It is demonstrated that the behaviour of the hydrodynamic field in the vicinity of the TE can be predicted with reasonable accuracy using triple deck theory if the eddy viscosity is accounted for. Point spectra of surface pressure difference are shown to vary considerably towards the trailing edge, with a significant reduction of amplitude occurring in the low frequency range.The acoustic pressure obtained from the DNS is compared with predictions from two- and three-dimensional acoustic analogies and the classical trailing edge theory of Amiet. For low frequencies, two dimensional theory succeeds in predicting the acoustic pressure in the far field with reasonable accuracy due to a significant spanwise coherence of the surface pressure difference and predominantly two dimensional sound radiation. For higher frequencies, however, the full three dimensional theory is required for an accurate prediction of the acoustic far field. DNS data are used to test some of the key assumptions invoked by Amiet for the derivation of the classical trailing edge theory. Even though most of the approximations are shown to be reasonable, they collectively lead to a deviation from the DNS results, in particular for higher frequencies. Moreover, because the three dimensional acoustic analogy does not provide significantly improved results, it is suggested that some of the discrepancies can be attributed to the approach of evaluating the far field sound using a Kirchhoff-type integration of the surface pressure difference
Going Beyond Counting First Authors in Author Co-citation Analysis
No full textThe present study examines one of the fundamental aspects of author co-citation analysis (ACA) - the way co-citation
counts are defined. Co-citation counting provides the data on which all subsequent statistical analyses and mappings
are based, and we compare ACA results based on two different types of co-citation counting - the traditional type that
only counts the first one among a cited work's authors on the one hand and a non-traditional type that takes into
account the first 5 authors of a cited work on the other hand. Results indicate that the picture produced through this non-traditional author co-citation counting contains more coherent author groups and is therefore considerably clearer. However, this picture represents fewer specialties in the research field being studied than that produced through the traditional first-author co-citation counting when the same number of top-ranked authors is selected and analyzed. Reasons for these effects are discussed
A comparative study of PtCo/C Alloys and Pt/C as cathode catalysts for fuel cell applications
No full textCommercialisation of fuel cells for automotive applications requires catalysts with both improved activity, particularly for the cathodic oxygen reduction reaction (ORR), and stability over commercial platinum only catalysts. This requires a detailed understanding of both desirable catalyst structure and mechanistic activity. In this work a comparative study is made between commercial Pt/C catalysts and developmental PtCo/C catalysts, with the aim of increasing understanding of the desirable characteristics of improved activity and stability demonstrated for PtCo catalysts.To explore the origins of activity and stability enhancement for Pt3Co/C catalysts over Pt/C, the mechanisms of catalyst deactivation have been investigated. In addition, four new PtCo3/C catalyst materials have been prepared and evaluated for ORR activity before and after activation by electrochemical dealloying. Structural changes occurring during these processes have been characterised using cyclic voltammetry, Transmission Electron Microscopy (TEM) and X-ray absorption spectroscopy (XAS).XAS has also been used to explore the relationship between catalyst activity and structure. This has been achieved by building and validating a specially adapted fuel cell, enabling study of fuel cell catalysts in situ under realistic operating conditions. Findings from time resolved potential step experiments with 12 s and 0.1 s time resolution will be presented for catalysts operating at OCV and under load with oxygen cathode gas. In addition, studies have been conducted with nitrogen cathode feed gas to investigate the effect of potential and temperature on the processes of oxide formation and removal on both Pt/C and PtCo/C catalysts
Edge-connected pixel task.
No full text(A) Illustration of the general connectivity task and its global nature. (B) Example of an input image and correct labelling for that image in the edge-connected pixel task. (C) Implementation of the single-hidden layer architecture for detecting whether the center pixel is connected to the edge. Each hidden neuron checks whether a certain pattern connecting the pixel to the edge is in the image. (D) Tag propagation solution in a recurrent network for detecting whether the center pixel is connected to the edge. We start with any on pixels connected to the edge tagged as connected. At subsequent time steps, the tag is passed to any neighboring pixel which is also on. (E) Schematic of the setup for the analytical solution of implementation of the tag propagation algorithm in neural network weights. The same setup is repeated at all pixels in the image. (F) Output of the analytical tag-propagation network at progressive time steps for several example images.</p
CPU, SMP and GPU implementations of Nohalo level 1, a fast co-convex antialiasing image resampler
No full textThis article introduces Nohalo level 1 (“Nohalo”), the simplest member of a family of image resamplers which straighten diagonal interfaces without adding noticeable nonlinear artifacts. Nohalo is interpolatory, co-monotone, coconvex, antialiasing, local average preserving, continuous, and exact on linears. Like many edge-enhancing methods, Nohalo has two main stages: first, nonlinear interpolation is used to create a double-density version of the original image; this doubledensity image is then resampled with bilinear interpolation. Nohalo is especially suited for GPU computing because the nonlinear slopes can be computed once and stored in a low bit-depth texture without rounding error, because the final bilinear stage can be performed in hardware, and because monotonicity allows full use of the texture’s dynamic range. Demand-driven implementations for CPUs and SMPs are more complex, and require extra work to fix bottlenecks. Efficient implementations of the minmod function are key to performance. Three implementations of Nohalo are presented and benchmarked: a CPU version in C for the graphics library GEGL, an SMP version in C++ for the graphics library VIPS and a GPU version in HLSL for DirectX. The GPU implementation is branch-free thanks to the discovery of a simple formula for the pixel values of the double density image. Branches are eliminated in the demand-driven C/C++ implementations by reflecting, if needed, Nohalo’s 12-point stencil with pointer shifts. Overall, Nohalo is not much slower than standard bicubic resamplers. Compared to twenty-three alternatives in tests involving the re-enlargement of images downsampled with nearest neighbour, Nohalo gets the best PSNRs
Memorandum from A. E. Demaray to E. C. Finney
No full textFour letters of correspondence about the purchase of Bright Angel Trail between A. E. Demaray, Acting Director of the Grand Canyon National Park; E. C. Finney, Department of the Interior First Assistant Secretary; Carl T. Hayden, Representative (AZ); and Stephen T. Mather, Director of the National Park Service
A morphological gradient approach to color edge detection
No full textA new color edge detector based on vector differences is proposed. The basic technique gives as its output the maximum distance between the vectors within a mask. When applied to scalar-valued images, the method reduces to the classic morphological gradient. The technique is relatively computationally efficient and can also be readily applied to other vector-valued images. To improve the performance in the presence of noise, a novel pairwise outlier rejection scheme is employed. A quantitative evaluation using Pratt's figure of merit shows the new technique to outperform other recently proposed color edge detectors. In addition, application to real images demonstrates the approach to be highly effective despite its low complexity
- …
