msaliency

说明：  图像分割，还可以，kmeans聚类的方法。k-means算法是一种聚类算法，所谓聚类，即根据相似性原则，将具有较高相似度的数据对象划分至同一类簇，将具有较高相异度的数据对象划分至不同类簇。(K-means algorithm is a clustering algorithm. The so-called clustering is to divide data objects with high similarity into the same class cluster and data objects with high dissimilarity into different class clusters according to the similarity principle.)

说明：  灰度图像二值化及目标提取和质心求取,Matlab代码.对初学者学习有用(Gray image binarization and target extraction and center of mass strike, Matlab code. Learn useful for beginners)

像素与图像基本指标—彩色图像的基本信息、像素颜色值、图像的大小、矩阵形式等、对灰度图像计算有关统计参数：（1）图像的大小；（2）图像的灰度平均值；（3）图像的自协方差；（4）图像的灰度标准差、对图像进行45度，19度，98度等任意角度的旋转，并计算原图与旋转后图像的相关系数和协方差矩阵(Pixel and Basic Index of Image)

利用提升小波方法对小波去噪，效果要比一般的方法好。(Use of lifting wavelet method of wavelet denoising, the method more effective than the general good.)

软岩隧道围岩稳定性和位移影响分析，研究围岩稳定性和位移的一篇论文，感觉挺好。(Soft Rock Tunnel stability and deformation analysis, stability and displacement of rock paper, feeling very good.)

说明：  利用BOW结合空间金字塔模型，使用支持向量机进行图像分类(Image classification using bow)

color image watermarking with color

说明：  图像处理软件含帮助。毕业设计，旋转，线性变换，灰化，平滑，滤镜，反相等，功能全面！(image processing software with help. Graduation design, rotary, linear transformation, ash, smooth, filter, anti-equal, full functions!)

TV模型用于图像修复，能够实现较好的图像修复效果，是基于偏微分方程的方法(TV model for image restoration, and to achieve better image restoration effect is a method based on partial differential equations)

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)