1,721,354 research outputs found
Determination of the shape of objects in the range of 1/20 of the wavelength of light
No full textClassical optics has a well defined limit of resolution that is given by λ/2. In the past this quantity was considered as a firm barrier that one could not overcome. Hence the solution to get higher resolutions was to use the particle-wave duality to get shorter wavelengths, for example the electron microscope. However, in 1952 Toraldo di Francia foresaw that the classical limit was in reality not as formidable as its looked at that time. He suggested ways to achieve higher resolutions (the multiple corona approach). Furthermore, he saw the connection between super resolution and evanescent light fields. He performed the first experiments showing the existence of evanescent fields in the range of visible optics In recent times a variety of experimental set ups have been utilized to go beyond the classical optics limit, the more effective ones based on the use of optical fibers tips to sense near fields (near field microscopy). In this paper, it is shown for the first time that full field near field observation is possible. Further by combining the optical advantages gained with the use of evanescent fields with numerical techniques of super resolution one can go well beyond the classical limit getting to the region of λ/8 by optical means and reaching the λ/20 with super resolution numerical techniques. Examples are shown of the observation of the shapes of small particles including some nanocrystals geometrie
An efficient Sequential Linear Programming algorithm for engineering optimization
No full textThis paper discusses the features of a new tool for engineering optimization problems. The new tool, Linearization Error Sequential Linear Programming (LESLP), includes an improved formulation of the well-established Sequential Linear Programming method. The LESLP algorithm is tested in seven structural optimization problems and two mathematical programming problems. The numerical efficiency, merits and limitations of LESLP are investigated in detail and compared with other general purpose optimization algorithms. Results indicate that LESLP is competitive with respect to other optimization algorithms published in the literature
Regularity results for a free interface problem with Holder coefficients
No full textWe study a class of variational problems involving both bulk and interface energies. The bulk energy is of Dirichlet type albeit of very general form allowing the dependence from the unknown variable u and the position x. We employ the regularity theory of A-minimizers to study the regularity of the free interface. The hallmark of the paper is the mild regularity assumption concerning the dependence of the coefficients with respect to x and u that is of Holder type
A numerical code for lay-out optimization of skeletal structures with Sequential Linear Programming
No full textThis paper discusses a new Sequential Linear Programming (SLP) algorithm denoted as LESLP (Linearization Error Sequential Linear Programming). The new algorithm implements an advanced strategy to choose the move limits which are defined by limiting the difference between the original nonlinear problem and its linearized counterpart. Besides, LESLP includes a trust region model to increase the design freedom and to improve the overall efficiency of the optimization process. LESLP is tested in five lay-out optimization problems of skeletal structures (i.e. bar trusses and frames) where the, objective is to minimize the weight of the structure. It is obviously intended that the objective function does not require structural analysis. The new algorithm is compared to other SLP based techniques and globally convergent optimization methods like Sequential Quadratic Programming (SQP). Results indicate that LESLP is competitive with recently published algorithms and commercial softwares. Also, the new algorithm is robust with respect to starting designs
Handy: A real-time three color glove-based gesture recognizer with learning vector quantization
No full textThis paper presents Handy, a real-time hand gesture recognizer based on a three color glove. The recognizer is formed by three modules. The first module, fed by the frame acquired by a webcam, identifies the hand image in the scene. The second module, a feature extractor, represents the image by a nine-dimensional feature vector. The third module, the classifier, is performed by means of learning vector quantization. The recognizer, tested on a dataset of 907 hand gestures, has shown very high recognition rate
- …
