200912260016

根据RBM深度网络（Hinton，2006）进行MINST手写文字的分类器的训练，利用训练得到的权值制作了这个小程序，通过这个程序可以看出训练结果对数据集内的测试样本和训练样本都能进行很好的识别，但是对其他的手写字体识别就没有那么好了。(According to RBM depth network (Hinton, 2006) conducted MINST handwritten text classifier training, using the training to get the right value to produce this small program, this program can be seen through the results of the training data set and the training sample test samples are can be well identified, but on the other handwriting recognition is not so good.)

EMD端点效应参考代码，包含多个算法：极值平均，边界特征尺度延拓，极值延拓，基于ISBM延拓，平行延拓，镜像延拓(EMD end effect reference code contains multiple algorithms: extreme average, the boundary characteristic scale extension, extreme extension, extension based ISBM parallel extension, mirror extension)

otsu算法应用于图像分割--图像分割-直方图(Otsu algorithm is applied to image segmentation image segmentation- Histogram)

说明：  模式识别与图像处理，图像的边缘处理，增强(Pattern recognition and image processing, image edge processing, to enhance)

实现了图像的灰度化并进一步用于视频监视控，一些自适应信号处理的算法，计算加权加速度。( Achieve a grayscale image and further control for video surveillance, Some adaptive signal processing algorithms, Weighted acceleration.)

说明：  压缩感知 ：非传统采样的模型实例和众多应用(model based compressivetoolbox)

这篇文字介绍了基于IBP重建的图像超分辨算法(this paper introduce an super-resolution Algorithm based on IBP)

运用Perona-Malik各项异性扩散模型对二维图像进行去噪处理。该方法在有效的抑制噪声的同时能够保护图像的边缘。(The use of Perona-Malik Anisotropic Diffusion Model for the treatment of two-dimensional image denoising. The method in the effective suppression of noise at the same time to protect the edge of the image.)

cashift_matlab,我做的一些修改： 1.修改了之前网上的那种需要输入avi文件名再读取的方式，改成直接读取avi文件。 2.修改了另外建立avi文件的方式。 2.实时显示跟踪结果。 3.彩色图显示，用红色框圈出目标。 4.因为是基于颜色特征的，所以要选色颜色特征明显的。你可以自己拍摄目标颜色突出的avi视频来测试。 5.程序结束时有点小问题，按ctrl+c结束。(cashift_matlab, I made some changes: 1. Modify the kind of network before you need to enter the avi file name and then read the way to read avi file directly. 2. Modify the other way to create avi file. 2. Real-time display tracking results. 3. The color map shows the target with a red box. 4. Because it is based on the color characteristics, so to choose the color characteristics of the obvious. You can test your own avi video with a prominent target color. 5. A little problem at the end of the program, press ctrl + c to end.)

使用sift+RANSAC完成两幅图像的特征提取和匹配，并将较小图像区域在另一幅中用方框圈出来。运行plot.m。(This code uses sift and RANSAC to extract features of two images and then finds and marks the smaller image in the other image.run plot.m)