transform

1 SIFT 发展历程 SIFT算法由D.G.Lowe 1999年提出，2004年完善总结。后来Y.Ke将其描述子部分用PCA代替直方图的方式，对其进行改进。 2 SIFT 主要思想 SIFT算法是一种提取局部特征的算法，在尺度空间寻找极值点，提取位置，尺度，旋转不变量。 3 SIFT算法的主要特点： a) SIFT特征是图像的局部特征，其对旋转、尺度缩放、亮度变化保持不变性，对视角变化、仿射变换、噪声也保持一定程度的稳定性。 b) 独特性(Distinctiveness)好，信息量丰富，适用于在海量特征数据库中进行快速、准确的匹配[23]。 c) 多量性，即使少数的几个物体也可以产生大量SIFT特征向量。 d) 高速性，经优化的SIFT匹配算法甚至可以达到实时的要求。 e) 可扩展性，可以很方便的与其他形式的特征向量进行联合。 4 SIFT算法步骤： 1） 检测尺度空间极值点 2） 精确定位极值点 3） 为每个关键点指定方向参数 4） 关键点描述子的生成 本包内容为sift算法matlab源码(1 SIFT course of development SIFT algorithm by DGLowe in 1999, the perfect summary of 2004. Later Y.Ke its description of the sub-part of the histogram with PCA instead of its improvement. 2 the SIFT main idea The SIFT algorithm is an algorithm to extract local features in scale space to find the extreme point of the extraction location, scale, rotation invariant. 3 the main features of the SIFT algorithm: a) SIFT feature is the local characteristics of the image, zoom, rotate, scale, brightness change to maintain invariance, the perspective changes, affine transformation, the noise also maintain a certain degree of stability. b) unique (Distinctiveness), informative, and mass characteristics database for fast, accurate matching [23]. c) large amounts, even if a handful of objects can also produce a large number of SIFT feature vectors. d) high-speed and optimized SIFT matching algorithm can even achieve real-time requirements. e) The scalability can be very convenient fe)

Matlab验证码识别算法实现，详细见代码实现，要搭建Matlab环境(Matlab verification code identification algorithm, see the code implementation in detail, to build Matlab environment)

Matlab code for supply ShipX from Hydro

说明：  利用Hu矩实现视频或是图像的特征提取，达到了很好的图像或是视频识别效果(Using Hu moments of video or image feature extraction, to achieve a good image or video recognition effect)

最短路径dijkstra算法的MATLAB实现(MATLAB implementation of the shortest path Dijkstra algorithm)

有关matlab的图像处理课件，供初学者学习参考。(Matlab image processing related courseware for beginners to learn information.)

潮流计算程序,用Ｍ语言实现简单电力系统潮流计算，已通过编译(powerflow)

变步长自适应滤波算法，对VS—LMS算法进行了改进，建立了步长因子μ与误差信号e（n）之间另一种新的非线性函数关系．理论分析和计算机仿真结果表明，该关系不仅具有原有算法收敛速度快的优点，而且在低信噪比环境下比原有算法具有更好的抗噪声性能．(The variable step size adaptive filtering algorithm is discussed, and the VS-LMS algorithm is improved. A new nonlinear function relation between the step size factor μ and the error signal e (n) is established. Theoretical analysis and computer simulation results show that the proposed algorithm not only has the advantages of high convergence speed of the original algorithm, but also has better anti- noise performance in the low SNR environment than the original algorithm.)

人脸识别匹配的源代码，并配有部分人脸照片，欢迎下载学习(The source code of Face recognition , with some face photos, welcome to learn.)

说明：  模糊建模工具箱，使用模糊聚类方法对系统进行辨识(Fuzzy modeling toolbox)