brox_zip

2014年西安电子科技大学数字图像处理上机实验，同时是研究生和本科生的代码及作业 (2014 Xi' an University of Electronic Science and Technology of digital image processing experiments on board, while the graduate and undergraduate students and job code)

说明：  数字图像处理常用程序，包括图像读取、显示、详细信息，对图像文件格式转换，RGB分量显示，索引图像、伪彩色图像、真彩色图像显示，灰度直方图处理，图像灰度变换，均衡化处理，图像加噪声处理，均值滤波、中值滤波，图像加减乘除运算，图像放大，插值拟合，旋转，多种边缘检测方法，高通滤波，低通滤波，带通滤波，维纳滤波，运动模糊等处理。(Common digital image processing programs, including image reading, display, detailed information, image file format conversion, RGB component display, index image, pseudo color image, true color image display, gray histogram processing, image gray transformation, equalization processing, image plus noise processing, mean filtering, median filtering, image plus, subtraction, multiplication and division, image amplification, interpolation Fitting, rotation, multiple edge detection methods, high pass filtering, low-pass filtering, band-pass filtering, Wiener filtering, motion blur and other processing.)

老外写的程序，可对三角形进行渲染，有详细的算法介绍。非常棒(A good Triangle Renderer Project，with detailed introduction.)

说明：  双目匹配计算视差，生成深度图。（前提是用自己标定好的摄像头拍的图片，导入内外参数然后进行立体匹配）(Binocular matching calculates parallax and generates depth map.(The premise is to take a picture with your own calibrated camera, import internal and external parameters, and then stereo matching))

极化SAR图像处理，经典的Lee滤波处理，可实现sar图像的去除斑点噪声处理(SAR image filtering based on Lee s algorithm.)

中国古代书法作品由于自然风化和人为因素导致的背景中包含大量噪声，重建背景图像，提高观赏价值。直接运行BackgroundFix.m即可(Since the background of ancient Chinese calligraphy works of natural weathering and human factors included a lot of noise, the background image reconstruction, improve the ornamental value. It can run directly BackgroundFix.m)

一些基本的图像处理算法的matlab实现，包括边缘提取和roc分析等(some basic image processing algorithms to achieve Matlab, including edge detection and analysis roc)

有关图像的目标识别："给出一种基于特征分类辨识的合成孔径雷达图像目标检测方法#用恒虚警和扩展分形方法对3&E图像进行目 标检测后用面积和峰值能量比算子辨识目标和背景杂波!去除一部分虚警!用小波域主成分分析对每个检测窗口内的图 像提取特征向量!用支持向量机对提取得到的特征向量进行分类!辨识目标和背景杂波!完成目标检测#使用&K?3数 据对该方法进行验证和分析!实验结果表明!经过特征分类辨识后!在检测率不变的情况下!虚警数目显著降低# (Related to the image target recognition)

针对杂波环境下的多个机动目标跟踪问题, 本文将多模型概率假设密度 (Multiple-model probability hypothesis density, MM-PHD) 滤波器和平滑算法相结合, 提出了 MM-PHD 前向 – 后向平滑器. 为了避免引入复杂的随机有限集 (Random finite set, RFS) 理论, 本文根据 PHD 的物理空间 (Physical space) 描述法推导得到了 MM-PHD 平滑器的后向更 新公式. 由于 MM-PHD 前向–后向平滑器的递推公式中包含有多个积分(By integrating the multiple-model probability hypothesis density (MM-PHD) filter with the smoothing al- gorithms, an MM-PHD forward-backward smoother is proposed in this paper for tracking multiple maneuvering targets in clutter. To avoid use of complex random finite set (RFS) theory, the backward updated equation of the MM-PHD smoother can be derived according to the physical-space explanation of the PHD)

说明：  利用Hessian矩阵的滤波函数Frangi，再血管图像处理上有很好的应用(Using the filter function Frangi of Hessian matrix, there is a good application in revascular image processing.)