iterative phase retrieval algorithms

红外图像的亚像素边缘检测，检测边缘，用于图像配准。(Infrared image of the sub-pixel edge detection, edge detection for image registration.)

说明：  卷积神经网络（Convolutional Neural Network）是一个专门针对图像识别问题设计的神经网络。它模仿人类识别图像的多层过程：瞳孔摄入像素；大脑皮层某些细胞初步处理，发现形状边缘、方向；抽象判定形状（如圆形、方形）；进一步抽象判定（如判断物体是气球）。(Convolutional neural network is a neural network specially designed for image recognition. It imitates the multi-layer process of human recognition image: pupil ingestion of pixels; some cells in the cerebral cortex preliminary processing to find shape edges and directions; Abstract judgment of shape (such as circles and squares); and further Abstract judgment (such as judging objects are balloons).)

本程序集合了互信息配准，PV差值，powell算法，基于小波变换的图像融合的一整套算法。只需要在GUI界面输入两幅图像，即可自动完成图像的配准融合。程序包含Matlab源代码。用Matlab打开imagefusion.m，即会弹出gui界面。融合前要输入参考和浮动图像。图像要求像素的 长＊宽 一样。否则会报错。程序只能处理灰度图像，彩色图像也会转化成灰度图像处理。程序代码已经有详细的注释，方便读者理解。(The procedures set of mutual information, PV difference, powell algorithm, based on wavelet transform image fusion set of algorithms. Only need to enter the two images in the GUI interface, you can automatically complete the image registration fusion. Program contains Matlab source code. Using Matlab open imagefusion.m, gui interface will pop up. Fusion before the input reference and floating images. Image requires the same pixel length* width. Otherwise incorrect. Can only handle grayscale images, color images into gray image processing also. Program code has detailed notes, to facilitate understanding.)

HOG LBP 在python matlab C++环境下的实现(Implementation of HOG LBP in Python matlab C++ environment)

易语言动画地图编辑器源码，易语言写的图形图像编程，很好的参考。(Easy language animated map editor source code, easy graphic language programming, a good reference.)

本代码主要是harris角点提取，ncc算法进行粗匹配，然后ransac算法剔除误匹配点(harris ncc ransac)

用matlab编写的实现椭圆检测的程序，自己稍微改下就可以(using Matlab prepared by the realization ellipse detection procedures, their own little can be changed under)

控制测量是测绘专业的基础课程，这个课件是武汉大学老师编的课件，很实用(Control measurements are mapping the professional foundation courses, this courseware is Wuhan University teachers series courseware, very practical)

图像拼接源码实现，包括多频率图像融合，sift特征点配准算法(Image Stitching source implementation, including multi-frequency image fusion, sift feature points matching algorithm)

多个函数的曲线拟合，里面统计有最小二乘、非线性回归(Curve fitting for multiple functions)