搜索资源列表
PSF
- 本文针对灰度图像的盲复原问题进行了一些研究,介绍了两种不同的图像盲复 原处理的方法。一种是误差一参数分析法,该方法适合于辨识可以用参数来表征的 点扩散函数,如线性移动模型和Gauss模型等,然后根据估计的参数确定退化图像 的点扩’散函数,再利用常规的复原算法(如维纳滤波法)对退化图像进行复原处理 另一种是非负支撑域约束递归逆滤波(NAS-R工F)算法,文中介绍了NAS-R工F算法 的基本思想,并结合正则化的思想,提出了NAS-RIF的改进算法,相应对该算法的 性能效果
bayes_bpnet
- 采用贝叶斯正则化算法提高 BP 网络的推广能力。我们采用两种训练方法,即 L-M 优化算法(trainlm)
Example4
- 采用贝叶斯正则化算法(抑制过拟合)提高 BP 网络的推广能力,采用两种训练方法, 即 L-M 优化算法(trainlm)和贝叶斯正则化算法(trainbr),用以训练 BP 网络;-Bayesian regularization algorithm (inhibition of over-fitting) to improve the generalization ability of BP network, using two training methods, that LM opti
corner
- 该程序找到了使用L曲线算法的拐点,是正则化程序中,寻求正则化参数的一种比较好的方法。CORNER Find corner of discrete L-curve via adaptive pruning algorithm.-The program found the algorithm using the L-curve inflection point, is the regularization process, the regularization parameter for a bet
bys
- 采用贝叶斯正则化算法提高BP网络的推广能力。在本例中,将采用两种训练方法,即L-M优化算法(trainlm)和贝叶斯正则化算法(trainbr),用以训练BP网络,使其能够拟合某一附加有白噪声的正弦样本数据。-The use of Bayesian regularization algorithm for BP network to improve generalization ability. In this case, two types of training methods will b
Bolasso-feature-selection-prediction
- 这个程序实现了Francis R. Bach的Bolasso算法,用于特征选取和预测。主要用于高纬度问题的特征选取,它使用了带有Bootstrap方法的自助抽样的正则化回归,并使用了Karl Skoglund的lars实现。-This procedure achieved Francis R. Bach s Bolasso algorithms for feature selection and forecasting. The main problem for high-latitude fe
trainlm
- 采用两种训练方法,即 L-M 优化算法(trainlm)和贝叶斯正则化算法(trainbr)-Using two training methods, namely, LM optimization algorithm (trainlm) and Bayesian regularization algorithm (trainbr)
regu
- 正则化化技术处理,l_curve方法和gcv方法选择参数,进行求解,得到问题的最解-Regularization of technical processing, l_curve methods and gcv method to select parameters, is solved by the solutions of the problem
1
- 正则化算法工具箱 多种方法实现 有相关注释文件-regulation tools
tikl_diff1
- 应用吉洪诺夫正则化的方法去求数值微分的求导问题-failed to translate
optimization
- 图像分割matlab 代码,不使用传统的tv 正则项,而是提出了一种新的正则化方法,可以运行-image segmentation with new regularization term
MFTV
- 图像超分辨率程序,基于稀疏表示和正则化方法,程序带有注释,提供相关论文. 可以直接运行. -Matlab code about super-resolution, based on sparse representation and regularization. The codes include annotation. Related paper also provided
predict-and-match-interal-multiple
- 地震信号处理,虚同相轴方法预测层间多次波,将数据分成上下两部分,利用相关和褶积的原理预测出层间多次波。预测信号和原始信号在相位和振幅上存在差异,用L1范数匹配法进行匹配,其中,提供了两种方法解病态方程,分别为高斯-赛德尔方法和正则化方法。-Seismic signal processing, predicte internal multiples by construct virtual events .The data is divided into two parts, using the
auto_inversion
- 求解第一类fedholm方程,运用philips光滑化方法和正则化方法求解。-Solving equations of the first kind fedholm using philips smoothing method and regularization method to solve.
LowPatchRank_regularization
- 对于带噪声的周期性纹理图像,提出一种基于二维秩约束的混合正则化去噪方法。该方法结合了全变分去噪理论和方法,并且利用该类图像低块秩的特性,对图像进行了低块秩约束。通过和全变分去噪方法比较可知,对于周期性纹理图像,混合正则化方法能有效地分离出噪声,并且能让图像很好地保持边缘。即使非严格的周期性纹理,该方法依然有很好的去噪效果。-For periodic texture images with noise, a new method based on two dimensional rank cons
Tikhonov-for-inverse-problem
- 《不适定问题的正则化方法及应用》pp.6例题1.2的Tikhonov正则化方法及迭代Tikhonov正则化方法程序包。-Tikhonov regularization method and iterative Tikhonov regularization method package for the book regularization method and applications of ill-posed problems pp.6 example 1.2.
csvd+ttls
- ttls的正则化方法,适用于各种病态反问题的求解(The regularization method of ttls is applicable to solving various ill conditioned inverse problems)
tgsvd+tsvd
- tgsvd正则化方法,适用于计算各类病态反问题(The tgsvd regularization method is suitable for the calculation of ill posed inverse problems)
tikhonov
- 解病态方程的正则化方法、逆问题不适定解的处理与正则化(Regularization method for solving ill conditioned equations)
cvpr16_deblur_study-master
- 文献 "Deblurring Text Images via L0-Regularized Intensity and Gradient Prior" 的参考代码 用Lp正则化方法做盲复原的代码 demo_text_deblurring 是主函数(refer to "Deblurring Text Images via L0-Regularized Intensity and Gradient Prior" main function: demo_t