排序方式: 共有113条查询结果,搜索用时 0 毫秒
51.
对称延拓小波变换矩阵用于FRIT去噪 总被引:2,自引:0,他引:2
给出了对称延拓方式下有限长信号不需逐级计算而直接得到小波系数的分解矩阵和由这些小波系数重构原信号的重构矩阵的构造方法,并给出了常用的相应于9/7小波的分解矩阵和重构矩阵及其基向量,它们可广泛用于基于小波的图像分块处理中.作为一种应用实例,将构造的小波变换矩阵用于FRIT图像去噪,不仅计算大大简化,而且相对于周期延拓的小波变换而言边界效应明显减少. 相似文献
52.
53.
基于脊波和神经网络的大压缩比遥感图像压缩 总被引:1,自引:0,他引:1
为了实现大压缩比的遥感图像压缩,利用神经网络的自组织、并行计算和分布式存储的能力,提出一种基于神经网络的压缩方法.在传统单隐层前向神经网络的基础上,该网络使用一种新的能有效处理直线型和曲线型奇异性的多尺度几何分析工具-脊波,作为隐层神经元的激活函数.它不仅具有神经网络压缩的优点;并且由于脊波良好的时、频和方向局域化特性,能够对遥感图像的边缘和轮廓实现更加有效的表示.仿真结果表明:该方法不仅能实现较高的压缩比,而且具有重建图像质量好、学习快速和鲁棒性强等优点. 相似文献
54.
针对传统数字水印算法抵抗旋转、缩放等几何攻击能力较弱的问题,提出一种基于尺度不变特征的脊波域量化水印算法.利用尺度不变特征变换构造稳定的图像局部区域,对局部区域实施脊波变换,选择能量较大方向上的脊波细节系数,利用量化索引调制策略嵌入水印信息.仿真结果表明,该算法具有较好的检测性能,在抗压缩、叠加噪声、缩放、剪切等攻击方面具有较强的鲁棒性. 相似文献
55.
基于炮集地震记录面波的高振幅、低频率和低速度特性及其与反射波在小波域和脊波域中分布区域不同,提出了一种衰减强干扰面波的联合一维连续小波变换和脊波变换方法。首先根据面波的高振幅和低频率特性,利用一维连续小波变换的时频局部化性质,对强干扰面波进行衰减;然后利用脊波变换将地震记录变换到时间、尺度和射线参数的三维空间;最后在脊波域对剩余干扰进行衰减。合成记录和实际资料处理的应用表明该方法不仅可以较好地衰减地震记录中的干扰面波,而且可以保护有效信号的低频成分,为地震资料后续处理和分析提供更好的保证。 相似文献
56.
针对Gurvelet变换采用的金字塔分解对图像细节表现的不足,我们提出利用全变差数字滤波器提取图像细节,然后对其采用基于分数阶傅立叶变换和投影-切片定理的Ridgelet变换,在变换域中由极小化极大误差准则进行阈值估计并对变换域系数进行阈值处理,以实现图像去噪.与金字塔分解相比,全变差数字滤波器能够简化图像分解并得到包含几乎所有细节的单幅图像,从而更有利于在Ridgelet域中进行降噪处理.实验结果表明,相对于Ridgelet和Curvelet变换的去噪方法,本文方法在抑制噪声的同时具有更有效的边缘保护能力,同时消除了边缘处的振荡,并且相对于Curvelet变换节省了计算. 相似文献
57.
58.
Shuyuan Yang Licheng Jiao Min Wang 《Frontiers of Electrical and Electronic Engineering in China》2008,3(2):198-203
In this paper, we propose a new directional multi-resolution ridgelet network (DMRN) based on the ridgelet frame theory, which
uses the ridgelet as the activation function in a hidden layer. For the multi-resolution properties of the ridgelet function
in the direction besides scale and position, DMRN has great capabilities in catching essential features of direction-rich
data. It proves to be able to approximate any multivariate function in a more stable and efficient way, and optimal in approximating
functions with spatial inhomogeneities. Besides, using binary ridgelet frame as the mathematical foundation in its construction,
DMRN is more flexible with a simple structure. The construction and the learning algorithm of DMRN are given. Its approximation
capacity and approximation rate are also analyzed in detail. Possibilities of applications to regression and recognition are
included to demonstrate its superiority to other methods and feasibility in practice. Both theory analysis and simulation
results prove its high efficiency.
__________
Translated from Journal of Xidian University, 2006, 33(4): 557–562 [译自: 西安电子科技大学学报(自然科学版)] 相似文献
59.
基于有限脊波变换的图像内容认证算法 总被引:1,自引:1,他引:0
在有限脊波变换(FRIT,finite ridgelet transform)域,提出了一种用于图像内容认证的半易损数字水印算法。提取图像在FRIT时产生的均值信息作为图像内容特征,并由它生成认证水印。利用Bayesian最小风险函数估计选择FRIT系数,把认证水印嵌入这些系数。实验结果表明,此算法简便高效,能很好地区分图像内容的恶意篡改和偶然修改,并能确切指出篡改的位置。 相似文献
60.
《International Journal of Hydrogen Energy》2022,47(55):23316-23323
In order to effectively predict fire risk of hydrogen pipeline leakage and improve the safe level of hydrogen pipeline, the finite Ridgelet neural network optimized by improved firefly algorithm is constructed. Related researches on pipeline leakage fire risk and corresponding prediction models are summarized. The finite Ridgelet neural network is constructed by using finite Ridgelet function as excitation function of node in hidden layer, and the structure of finite Ridgelet neural network is established. The improved firefly algorithm based on location update mechanism is designed to optimize the parameters of finite Ridgelet neural network. And then the fire risk level prediction procedure of hydrogen pipeline is designed. Finally, forty training samples and ten testing samples are selected to carry out fire risk level prediction analysis, results show that the proposed finite Ridgelet neural network optimized by improved firefly algorithm has advantages in prediction precision and accuracy, which can effectively predict the safety status of hydrogen pipeline. 相似文献