pgm7

维纳滤波器仿真，是对基本算法原理的实现，没有考虑误差模型(Wiener filter simulation, the realization of the basic algorithm principle, without considering the error model)

m_map使用方法详细介绍，采用matlab下载世界地图，可以获取海岸线等数据库(M_map uses the method to introduce in detail, uses the MATLAB to download the world map.)

本文主要是用matlab提供的小波包对图像进行分解、压缩、重构(This article is mainly used to provide matlab wavelet packet decomposition of images, compression, Reconstruction)

说明：  利用那些已经被破坏的区域的边缘， 即边缘的颜色和结构，根据这些图像留下的信息去推断被破坏的信息区的信息内容，然后对破坏区进行填补 ，以达到图像修补的目的(Using the edge of the damaged area, i.e. the color and structure of the edge, the information content of the damaged information area can be inferred from the information left by these images, and then the damaged area can be filled to achieve the purpose of image repair.)

本程序用于数字水印，可以输入一副图像，嵌入水印信息，然后防止各种各样的攻击，最后可以提取水印信息，恢复原图像(This procedure for digital watermarks, can enter a pair of images, embedded watermark information, and then to prevent the various attacks, the watermark information can be extracted Finally, restore the original image)

高斯计算程序，高斯计算程序，高斯计算程序，高斯计算程序，高斯计算程序，高斯计算程序，(Gaussian calculation program, Gaussian calculation program, Gaussian calculation program, Gaussian calculation program, Gaussian calculation program, Gaussian calculation program,)

多分辨率灰度和旋转不变纹理分类与定位二进制模式(multiresolution gray scale and rotation invariant texture classification with location binary patterns )

图像检测的算法，Faster R-CNN算法，先对整张图像进行卷积计算，然后通过感兴趣区域池化层(RoI Pooling Layer)将选择性搜索算法推荐出来的候选区域和卷积网络计算出的特征映射图进行融合，得到候选区域对应的特征矢量，这种共享卷积计算的操作极大地减少了卷积计算的次数。而且这些特征矢量的维度统一，方便后续的分类工作。通过感兴趣区域池化层处理卷积特征，并将得到的特征送往两个并行计算任务进行训练，分类和定位回归。通过这些方法和改进的框架，Fast R-CNN 用更短的训练和测试时长，取得了比 R-CNN 更好的效果(Faster R-CNN algorithm first convolutes the whole image, then fuses the candidate regions recommended by the selective search algorithm and the feature mapping maps calculated by the convolution network through the RoI Pooling Layer to get the corresponding feature vectors of the candidate regions, which greatly reduces the number of convolution calculations. Moreover, the dimension of these feature vectors is unified, which facilitates the subsequent classification work. The convolution feature is processed by the pooling layer of the region of interest, and the obtained feature is sent to two parallel computing tasks for training, classification and positioning regression. Through these methods and improved framework, Fast R-CNN uses shorter training and testing time and achieves better results than R-CNN.)

数字水印参考代码，分别用checksum，lsb，patchwork等方式嵌入水印，并且附带计算信噪比和消除水印的代码(Digital watermark reference code checksum lsb in patchwork way embedded watermark and accompanying calculated signal-to-noise ratio and eliminate the watermark code)

主要思想是用meanshift算法跟踪目标，拖动矩形框选择目标模板，meanshift自动跟踪 本程序的主要缺陷体现在帧差法自动识别目标，这部分功能与上一个功能独立，可用可不用，用于自动识别目标，效果不是很好，主要体现在对图像视频的要求是 背景（目标以外的所有事物）不能动(The main idea is to use meanshift algorithm to track the target and drag a rectangle to select the target template, meanshift automatic tracking The main drawback of the program is reflected in the frame difference method automatically identifies the target, which is part of the function and one function independently, may or may not, be used to automatically identify the target, the effect is not very good, mainly reflected in the requirements of the video image is the background (target All things outside) can not move)