BP

说明：  基于最小二乘法的仿射变换 matlab实现(Least square method based on the realization of affine transformation matlab)

LLE算法，matlab源码，LLE算法是流形学习算法的一种(LLE algorithm, matlab source, LLE algorithm manifold learning algorithm is a)

基于NCC与SSDA的快速特征点匹配融合算法（图像匹配）(NCC and the SSDA-based feature points matching the rapid integration algorithm (matching))

基于低频融合算法和小波变换编写的图像融合Matlab程序，包含了基本的小波图像分解与重构程序。(Fusion algorithm based on the low-frequency and wavelet transform image fusion prepared Matlab procedures, including the basic wavelet image decomposition and reconstruction process.)

说明：  根据：基于不变矩特征和神经网络的图像模式模糊分类 论文 在matlab上做的实验，有7个不变矩生成和神经网络分类的代码，还要论文原文(According to: Based on the characteristics of moment invariants and neural network image mode fuzzy classification papers in matlab to do the experiment, there are seven moment invariants and neural network classifier to generate the code, but also the original papers)

实现基本的LMS算法和归一化LMS算法学习曲线的比较仿真(LMS algorithm and normalized LMS algorithm learning curve comparing simulation)

Canny进行边缘检测，得到图像的边缘，然后sobel算子对图像边缘进行卷积，然后用Hough变换进行直线提取。(Canny edge detection to obtain the edge image, and then sobel edge operator for convolution, and then use the Hough transform for straight line extraction.)

一个用于电子显微镜下运动细胞的识别跟踪的源代码，出自哈佛医学院的一位教授，对于做物体运动识别估计的各位很有帮助！(Under an electron microscope for cell identification and tracking movement of the source code, from a professor at Harvard Medical School, for the movement of objects do you identify the estimated useful!)

free-vision library,计算机视觉C++库，含有SAMPLE,非常经典！(free-vision library, computer vision C library, containing SAMPLE. very classic!)

（1）计算图像中每个像素点的LBP模式(等价模式,或者旋转不变+等价模式)。 （2）然后计算每个cell的LBP特征值直方图，然后对该直方图进行归一化处理(每个cell中，对于每个bin,h[i]/=sum,sum就是一副图像中所有等价类的个数)。 （3）最后将得到的每个cell的统计直方图进行连接成为一个特征向量，也就是整幅图的LBP纹理特征向量； 然后便可利用SVM或者其他机器学习算法进行分类识别了。((1) calculate the LBP pattern of each pixel in the image (equivalent mode, or rotation invariant + equivalent mode). (2) then the LBP eigenvalue histogram of each cell is calculated, and then the histogram is normalized (for each cell, for each bin, h[i]/=sum, sum is the number of all the equivalent classes in a pair of images). (3) finally, the statistical histogram of each cell is connected into a feature vector, that is, the LBP texture feature vector of the whole picture. Then, SVM or other machine learning algorithms can be used for classification and recognition.)