new_SPIHT

SURF Feature is an algorithm that is used in the object recognition please see the source code

将两张图像进行图像相减或相处，使两张图片叠加。(The two images are subtracted or coexisted, so that the two images are superimposed.)

说明：  可通过Hog和svm实现小狮子的分类，将其分为两类(The classification of small lions can be realized by hogs and SVM, which can be divided into two categories)

提取GIST特征的matlab实现源码，初学者实用(Extract GIST features matlab achieve source beginners practical)

基于OPENCV的SIFT特征提取与匹配算法。包含完整的从图像高斯金字塔、DOG、空间极值点提取、关键点描述、KDtree匹配等关键步骤的全部函数实现，对全面深入理解Lowe的SIFT算法有莫大帮助。程序运行前须安装（1）OpenCV: http://opencvlibrary.sourceforge.net （2）SIFT: http://web.engr.oregonstate.edu/~hess/index.html，并配置其环境参数。(OPENCV the SIFT-based feature extraction and matching algorithm. Contains a complete Gaussian pyramid from the image, DOG, space extremum point extraction, description of key points, KDtree matching key step in the realization of the full function of the comprehensive and in-depth understanding of Lowe s SIFT algorithm of tremendous help. Program to run before the installation of (1) OpenCV: http://opencvlibrary.sourceforge.net (2) SIFT: http://web.engr.oregonstate.edu/ ~ hess/index.html, and configure the parameters of their environment.)

偏微分方程图像复原代码，其中有频域迭代盲解卷积H1模型，TV去噪模型，自己编的自适应去噪模型，以及改进的自适应去噪模型，令附卫星图片，PSNR代码(Image restoration code of partial differential equations, including iterative frequency-domain blind deconvolution H1 model, TV denoising model, made themselves adaptive denoising model, and improved adaptive denoising model, so attached to satellite images, PSNR code)

该代码，利用GDAL 1.4.2库实现遥感图像的几何校正、图像配准等，vs 2010实现。(The code GDAL 1.4.2 library implementation of remote sensing image geometric correction, image registration, vs 2010.)

说明：  对边缘图像作hough变换，从峰值中恢复出车道线(Right edge of the image for hough transform, recovered from a peak in the lane)

利用奇异值分解（SUV）对图像信号分行降噪！(Singular value decomposition (SUV) of the image signal branch noise!)

一种基于改进DRLSE的运动目标分割方法,为了得到精准、连续的运动目标轮廓，提出将帧间差分法和改进的距离规则水平集演化(DRLSE)方法结合应用于运动目标分割。(In order to get an accurate and continuous contour of moving target, a novel method to segment moving object by combining the frame difference and an improved distance regularized level set evolution (DRLSE) is proposed. )