首页 | 本学科首页   官方微博 | 高级检索  
相似文献
 共查询到20条相似文献,搜索用时 46 毫秒
1.
We study by means of -convergence the asymptotics of the rescaled Mumford-Shah functional
when and prove the existence of a -limit. The limit functional is easy to handle and can be used as a simple approximation to the original Mumford-Shah functional. Moreover, its minimizers can be interpreted as a sort of asymptotic probability distribution of the sets . Some examples illustrate the use of this method in image segmentation.Received: 12 June 2004, Accepted: 12 July 2004, Published online: 10 December 2004  相似文献   

2.
We present a generalized version of the Hardy-Sobolev inequality, in which the homogeneous potential is replaced by any potential V belonging to the Lorentz space . We show that the best constant in these inequalities is achieved provided that where . We also analyze the limit case . Finally an application to a non-linear eigenvalues problem with rough potentials is presented.Received: 19 September 2004, Accepted: 15 November 2004, Published online: 22 December 2004  相似文献   

3.
We consider the boundary value problem in a bounded, smooth domain in with homogeneous Dirichlet boundary conditions. Here 0,k(x) $$ " align="middle" border="0"> is a non-negative, not identically zero function. We find conditions under which there exists a solution which blows up at exactly m points as and satisfies . In particular, we find that if , 0 $" align="middle" border="0"> and is not simply connected then such a solution exists for any given Received: 11 February 2004, Accepted: 17 August 2004, Published online: 22 December 2004  相似文献   

4.
We prove the existence of continuously differentiable solutions such that
or
where
  相似文献   

5.
Estimates for deviations are established for a large class of linear methods of approximation of periodic functions by linear combinations of moduli of continuity of different orders. These estimates are sharp in the sense of constants in the uniform and integral metrics. In particular, the following assertion concerning approximation by splines is proved: Suppose that is odd, . Then
moreover, for it is impossible to decrease the constants on . Here, are some explicitly constructed constants, is the modulus of continuity of order r for the function f, and are explicitly constructed linear operators with the values in the space of periodic splines of degree of minimal defect with 2n equidistant interpolation points. This assertion implies the sharp Jackson-type inequality
. Bibliography: 17 titles.  相似文献   

6.
Consider the minimization problem
in which is a normal integrand. Define the convex function by It is known that, if the essential domain H of G is open, then problem (P) has a minimizer for any pair of endpoints (u 0, u 1). In this paper, the same result is proved under the condition that, for every point p in H, the subgradient set G(p) is either bounded or empty (when H is open, this condition holds automatically).  相似文献   

7.
Consider the convergence of the projection methods based on an extension of a special class of algorithms for the approximation--solvability of the following class of nonlinear quasivariational inequality (NQVI) problems: find an element such that and
where are mappings on H and K is a nonempty closed convex subset of a real Hilbert space H. The iterative procedure is characterized as a nonlinear quasivariational inequality: for any arbitrarily chosen initial point x 0 K and, for constants 0$$ " align="middle" border="0"> and 0$$ " align="middle" border="0"> , we have
where
This nonlinear quasivariational inequality type algorithm has an equivalent projection formula
where
for the projection P K of H onto K.  相似文献   

8.
Dedicated to Professor Jacque-Louis Lions on the occasion of his 70th birthday We consider a mixed problem for the operator
in a noncylindrical domain . We obtain local solution in t. When we add a viscosity we obtain a global solution. We also investigate the asymptotic behavior of the energy.  相似文献   

9.
For suitable positive integers n and k let m(n, k) denote the maximum number of edges in a graph of order n which has a unique k-factor. In 1964, Hetyei and in 1984, Hendry proved for even n and , respectively. Recently, Johann confirmed the following conjectures of Hendry: for and kn even and for n = 2kq, where q is a positive integer. In this paper we prove for and kn even, and we determine m(n, 3).  相似文献   

10.
In what follows, $C$ is the space of -periodic continuous functions; P is a seminorm defined on C, shift-invariant, and majorized by the uniform norm; is the mth modulus of continuity of a function f with step h and calculated with respect to P; , ( ), ,
,
Theorem 1. Let . Then
For some values of and seminorms related to best approximations by trigonometric polynomials and splines in the uniform and integral metrics, the inequalities are sharp. Bibliography: 6 titles.  相似文献   

11.
On the basis of the exact solution of the linear Dirichlet problem , we obtain conditions for the solvability of the corresponding Dirichlet problem for the quasilinear equation u ttu xx = f(x, t, u, u t).  相似文献   

12.
Let be the space of 2-periodic functions whose (r – 1)th-order derivative is absolutely continuous on any segment and rth-order derivative belongs to L p, S 2n,m is the space of 2-periodic splines of order m of minimal defect over the uniform partition . In this paper, we construct linear operators such that
where
To construct the operators X n,r,m, we use the same idea as in the polynomial case, i.e., the interpolation of Bernoulli kernels. As is proved, the operators X n,r,m converge to polynomial Akhiezer–Krein–Favard operators as . Bibliography: 10 titles.  相似文献   

13.
For entire Dirichlet series of the form , we establish conditions under which the relation
holds uniformly in outside a certain set E for which
where h() is a positive continuous function increasing to + on [0, +).  相似文献   

14.
The paper deals with the problem of recovering the parameters (functions) and of the Maxwell dynamical system
(tan is the tangent component; is a solution) by the response operator ( is the normal). The parameters determine the velocity , the c-metric , and the time . It is shown that for any fixed , the operator determines and in uniquely. Bibliography: 15 titles.  相似文献   

15.
Sufficient conditions of solvability and unique solvability of the boundary value problem are established, where are measurable functions and the vector function is measurable in the first and continuous in the last kmn arguments; moreover, this function may have nonintegrable singularities with respect to the first argument.  相似文献   

16.
The purpose of this paper is to give characterizations for uniform exponential dichotomy of evolution families on the real line. We consider a general class of Banach function spaces denoted and we prove that if with and the pair is admissible for an evolution family then is uniformly exponentially dichotomic. By an example we show that the admissibility of the pair for an evolution family is not a sufficient condition for uniform exponential dichotomy. As applications, we deduce necessary and sufficient conditions for uniform exponential dichotomy of evolution families in terms of the admissibility of the pairs and with   相似文献   

17.
We consider the subcritical problem & 0 & \qquad\textrm{in} \; A\\ u & = & 0 & \qquad\textrm{on} \; \delta A\\ \end{array} \right.$$" align="middle" border="0"> where A is an annulus in , , is the critical Sobolev exponent and 0$" align="middle" border="0"> is a small parameter. We prove that solutions of (I) which concentrate at one or two points are axially symmetric.Received: 7 July 2003, Accepted: 10 May 2004, Published online: 16 July 2004Filomena Pacella: Research supported by MIUR, project Variational Methods and Nonlinear Differential Equations.  相似文献   

18.
For f L n (T d ) and , the modulus of smoothness
is shown to be equivalent to
where T n is the best trigonometric polynomial approximant of degree n to f in L p and is the Laplacian. The above result is shown to be incorrect for 0 < p .  相似文献   

19.
We study the regularizing effect of perimeter penalties for a problem of optimal compliance in two dimensions. In particular, we consider minimizers of
where
The sets , , and the force f are given. We show that if we consider only scalar valued u and constant , or if we consider the elastic energy , then is away from where is pinned. In the scalar case, we also show that, for any of class , is . The proofs rely on a notion of weak outward curvature of , which we can bound without considering properties of the minimizing fields, together with a bootstrap argument.Received: 5 March 2002, Accepted: 3 September 2002, Published online: 17 December 2002  相似文献   

20.
We solve Tikhomirov's problem on the explicit computation of sharp constants in the Kolmogorov type inequalities
Specifically, we prove that
for all and k{0,...,n-1}. We establish symmetry and regularity properties of the numbers A n,k and study their asymptotic behavior as n for the cases k=O(n 2/3) and k/n(0,1).Similar problems were previously studied by Gabushin and Taikov.  相似文献   

设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号