WQ_matting_imerode

一种效果比较好的高光谱解混算法，模拟数据以及真实数据的实验都在，直接可以跑。(An effect good hyperspectral unmixing algorithm, experimental simulation data and real data are in, you can directly run.)

模糊聚类图像分割，还包括spectral clustering及其改进版（按照pr2010某paper）(image segmentation using fuzzy c means & spectral clustering)

图相多描述编码matlab仿真 多描述编码的matlab仿真，本科生的毕业设计 生成图像后计算psnr >> J=imread( 两描述都收到后重建的图像.bmp ) PSNR(I,J) ans = 32.3290 >> J=imread( 描述1重建后的图像.bmp ) >> PSNR(I,J) ans = 28.7011 >> J=imread( 描述2重建后的图像.bmp ) >> PSNR(I,J) ans = 29.1467 (Figure phase multiple description coding matlab simulation Matlab simulation of multiple description coding, undergraduate graduation design Generated images calculated psnr >> J = imread ( the reconstructed image after the two descriptions are received. Bmp ) >> PSNR (I, J) ans = 32.3290 >> J = imread ( to describe a reconstruction of images. Bmp ) >> PSNR (I, J) ans = 28.7011 >> J = imread ( to describe two reconstruction images. Bmp ) >> PSNR (I, J) ans = 29.1467)

该程序实现了相位编组法检测，可以将灰度缓慢变化的边缘提取出来(Realize the process of phase formation assay can be gray-scale changes in the edge of the slow extracted)

基于多尺度MRF模型的图像分割算法。使用提升小波对原始图像进行分解，使用ICM算法进行分割处理(Multi-scale MRF model based image segmentation algorithm. The original image using lifting wavelet decomposition, the use of ICM segmentation algorithm)

该程序是一种最小均方误差滤波算法的matlab实现，主要应用于干扰对消方面(The program is a minimum mean square error filtering algorithm matlab realize, mainly used in interference cancellation)

vssamp算法，图像变步长自动匹配重构重构算法，基于samp算法，更快速有效的重构复杂图像(vssamp variable step size image reconstruction algorithm automatically matches reconstruction, samp algorithm based on more quickly and efficiently reconstruct complex images)

基于粒子滤波和卡尔曼滤波的运动目标视频跟踪程序，非常强大和实用，强烈推荐学习使用。(Kalman filter based on particle filter and target video motion tracking program, very powerful and practical, it is strongly recommended to learn to use.)

计算PSNR值，一个不错的算法，值得看一下(Calculation of PSNR value, a good algorithm, it is worth a look)

说明：  一些快速实现图像模糊增强的的demo，matlab源码。已经测试通过。(Some image blurring rapidly enhanced the demo, matlab source. Have the test.)