figure4_41

SIFT 算法的实质上是提取局部特征的算法, 该算法在尺度空间寻找极值点, 然后根据极值点来获得尺度、旋转不变量和位置等特征。(In essence, the SIFT algorithm is an algorithm to extract local features. The algorithm finds the extreme points in the scale space and then obtains the characteristics of scale, rotation invariant and position according to the extreme point.)

很好的倒频谱计算程序，可立即生成图形，方便读者理解运用。(very good。)

说明：  基于局部Fisher准则的非线性核Fisher辨别分析，应用于有监督的特征提取与高维数据的有效降维。(Kernel Local Fisher Discriminant Analysis for Supervised Dimensionality Reduction.)

说明：  DOG高斯差分实现物体识别中的特征提取,对不同尺度空间的图像进行差分，从而获得描述物体的特征值(difference of gaussian to achieve the feature point exaction of object recognition)

目前比较流行的稀疏分解重构程序，可以用在人脸识别、字典构造等方面。(Currently popular sparse decomposition and reconstruction process, can be used in face recognition, dictionary structure and so on.)

能进行基于Arnold算法的正变换和反变换(Arnold-based algorithm to the transformation and anti-transform)

软阈值去噪加中值滤波，m文件，直接运行，有gui界面(Plus soft thresholding median filter, m files, directly run a gui interface)

说明：  图片采用3x3均值滤波，用Verilog语言描述，输入输出分别使用外同步(Pictures are filtered with 3x3 mean and described in Verilog language. Input and output are synchronized with each other.)

简单写的一些图像预处理（灰度增强、小波高通滤波、LBP、迭代阈值分割、种子生长法etc）和RGB 、纹理（相关矩阵）特征提取(Write some simple image pre-processing (grayscale enhancement, wavelet high-pass filter, LBP, iterative threshold segmentation, seed growth method etc) and RGB, texture (correlation matrix) feature extraction)

s变换；广义s变换；高斯邻域；时频分析；滤波(s transformation generalized s transformation Gaussian neighborhood time-frequency analysis Filtering)