Application of the Fundamental Solution Method to object recognition in the pictures

Recognition of objects in pictures and movies requires the use of techniques, such as filtering, segmentation and classification. Image filtering is required to remove all artifacts that hinder the unequivocal identification and sharpen interesting objects. Segmentation refers to finding areas of images respected to individual objects. For the selected areas corresponding to objects in the selected picture, the classification of objects finally gives information about the type of object which orientation is made. This paper presents a method for the classification of objects from drawings as a bitmap using the method of fundamental solutions (MFS). The MFS was tested on the selected bitmap depicting simple geometric shapes. The correlations between errors occurring on the boundary for particular shapes are used for the selection of geometric shape figures. Due to this correlation, it is possible to recognize the shape of the image appearing on the drawing by an analysis consisting of the comparison of recognized points describing the shape of contour to a database containing solutions of boundary value problems for the selected shape. In one way, the comparison of the pattern can determine which shape from database it is most similar to in terms of contour. This article appear that this approach is very simple and clearly. In result, this method can be used to recognition of the objects in the systems of real-time processing.

Fundamental Solution Method in a few MATLAB lines

We report a computational exposition to approximate solution of Laplaceequation on the plane using Fundamental Solution Method. For agiven boundary conditionsat the boundary, we recoverthe approximate solution in the region. We demonstrate thatfundamental solution method for the model leads to anunconstrained Quadratic Programming problem. To address the issue ofill-conditioning, we suggests to use SVD to overcome this issue.The implementation of the methods is using MATLAB.

Application of the Fundamental Solution Method to Object Recognition in the Pictures

Recognition of objects in pictures and movies requires the use of techniques, such as filtering, segmentation and classification. Image filtering is required to remove all artifacts that hinder the unequivocal identification and sharpen interesting objects. Segmentation refers to finding areas of images respected to individual objects. For the selected areas corresponding to objects in the selected picture, the classification of objects finally gives information about the type of object which orientation is made. This paper presents a method for the classification of objects from drawings as a bitmap using the method of fundamental solutions (MFS). The MFS was tested on the selected bitmap depicting simple geometric shapes. The correlations between errors occurring on the boundary for particular shapes are used for the selection of geometric shape figures. Due to this correlation, it is possible to recognize the shape of the image appearing on the drawing by an analysis consisting of the comparison of recognized points describing the shape of contour to a database containing solutions of boundary value problems for the selected shape. In one way, the comparison of the pattern can determine which shape from database it is most similar to in terms of contour. This article appear that this approach is very simple and clearly. In result, this method can be used to recognition of the objects in the systems of real-time processing.

Pareto optimization of radar receiver low-noise amplifier source impedance for low noise and high gain

In radar receivers, the low noise amplifier(LNA)must provide very low noise figure and high gain to successfully receive very low signals reflected off of illuminated targets. Obtaining low noise figure and high gain, unfortunately, is a well-known trade-off that has been carefully negotiated by design engineers for years. This paper presents a fundamental solution method for the source reflection coefficient providing the maximum available gain under a given noise figure constraint, and also for the lowest possible noise figure under a gain constraint. The design approach is based solely on the small-signal S-parameters and noise parameters of the device; no additional measurements or information are required. This method is demonstrated through examples. The results are expected to find application in design of LNAs and in real-time reconfigurable amplifiers for microwave communication and radar receivers.

Solutions for Triangularly Distributed Harmonic Loadings on Axially Symmetric Area in Layered Half-Space

ABSTRACTThe paper is to solve the problem of the response of stratified half-space subjected to triangularly distributed time-harmonic loadings on an axial symmetric area in the stratified half-space. Since the distributed area can be very small, the solution can be employed to deal with the problems of elastodynamics using the fundamental solution method. The advantage of the presented solution over Green function is that no singularity will occur in the presented solution. Some numerical results are given and some conclusions have been made. The presented solution is very efficient in computation.

The Usage of the Method of Fundamental Solution for Twisting Composite Rod as a Basis of the Local Analysis

The paper presents a resilient torsion of composite bars, which has a cross-section of a circle with a central circular fiber. Composite material is material with heterogeneous structure. It is composed of two or more components with different properties. The problem of torsion of homogeneous prismatic bars was discussed in many publications, however, there is very little work, concerning compute rods. In general form, the discussed question was first asked and solved mathematically by Мусхелишвили. He used a twisting function φ (X, Y), analogic to the stresses function and the rule of de Saint-Venant for homogeneous prismatic bars. On the basis of his work other were created, which include, Векуа i Рухадзе In this article the function of stress presented by Чобаня is used to solve the issue of composite torsion of prismatic rods. On this basis, it was found that the introduction of the stress function ψ (X, Y) simplifies the way of solving prismatic bars composed of different materials, especially when they have polygonal contours. The application of the theory of the function of twisting stresses in the composite rods allows the use approximate methods which can include: variational method, small parameter method or method used in the fundamental solution method. The analytical results were verified using numerical methods [1]. In the fundamental solution method stress concentrations occurring at the border of the fiber and the matrix was used to solve the problem of the composite rod twisting. This gives better results than the use of local analysis