1

本文主要是针对传统混合高斯模型运动目标检测算法的不足做出了一些改进(This article is for the shortcomings of traditional Gaussian mixture model moving object detection algorithm made some improvements)

基于Harris角点检测的图像拼接matlab实现， 主要内容包括Harris角点检测，NCC,RANSAC以及2幅图像的拼接(A image mosaicing algorithem which based on Harris corner feature detection achieved in matlab. The main contents include the Harris corner detector, the NCC, the RANSAC and two image mosaicing)

这本教科书，介绍了稀疏和多余的申述，对信号和图像处理应用的重点。的理论和数值基础处理前的应用进行了讨论。信号源的数学建模一起讨论如何使用适当的模型，如去噪，恢复，分离，插值和外推法，压缩，采样，分析和合成，检测，识别，多任务。这次报告会是优雅和迷人的。(Sparse and Redundant Representations From Theory to Applications in Signal and Image Processing )

批处理与重投影MODIS MOD04 数据(Reproject MODIS s mod04 data)

图像经过小波变换之后提取子带系数源代码在MATLAB中的实现(subbands in matlab)

说明：  dct-svd域水印算法,实现数字水印图像在载体DCT-SVD域的隐藏(a kind of watermarking algorithm in DCT-SVD domain)

本文主要介绍遗传算法的基本理论，叙述遗传算法在图像增强的的主要应用，即将原始图像变得更加清晰，特征变得更加明显。 现今关于图像增强的算法有很多，而这些算法大多是基于退化函数或者点扩展函数的知识进行图像处理的。当图像出现模糊或噪声影响大时，设计出的图像清晰化的效果肯定不够理想，因此有必要对图像进行增强处理。于是，可利用到遗传算法这种成熟稳定的仿生物进化的全局寻优算法，进行图像增强，由于遗传算法控制参数少、自适应度高，将选择该方法对图像退化分别进行不同的清晰化处理. 这样既增强了图像的对比度,又克服了传统直方图均衡化处理所造成的灰度级损失等缺点。 通过实验去表明，遗传算法从全局寻优的角度出发，能够较为精确地估计退化系统函数。和传统的线性增强、直方图均衡进行比较。实验结果表明该方法能改善原图像视觉效果，便于之后的图像分析。 (This paper mainly introduces the basic theory of genetic algorithm, genetic algorithm (ga) in the main application of image enhancement, the original image is more clear, characteristic becomes more obvious. Today about the image enhancement algorithm are many, and most of these algorithms is based on the degradation function or the knowledge of the point spread function of image processing. When the image appears fuzzy or noise influence, designed the image clarity of effect must not ideal, therefore it is necessary to enhance the image. Then, genetic algorithm is available to the mature and stable evolution of the global optimization algorithm, for image enhancement, due to less genetic algorithm control parameters, since the high fitness, will choose the method of image degradation, respectively, to manipulate the different motivation. It can enhance the image contrast, and overcome the traditional histogram equalization processing grayscale loss caused by the faults. Throug)

可以实现图像的重采样，通过设置参数来减小或增大图像尺寸(Resampling of the image can be achieved by setting parameters to decrease or increase the image size)

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

说明：  ORL人脸图像库，共40人，每人10幅图像，其中每人的前5幅作为训练样本，后5幅作为测试分类样本，统计正确分类率。分类准则为最近邻规则。 真实的图像尺寸为112x92，列向量堆积对应人脸库矩阵的每一列。 (ORL face image database, a total of 40 per 10 images, each of which the first five as training samples, after the 5 categories as a test sample, correct classification rate statistics. Classification criteria for the nearest neighbor rule. The real image size is 112x92, the corresponding column vector face database matrix accumulation of each column.)