JIAMI

图像的腐蚀、膨胀、开、闭运算。用菱形算子进行。(Image of erosion, dilation, opening and closing operation)

在边缘检测之后得到的边缘不一定是完整的，利用这个函数可以将个线段生长一些，然后连接起来，成完整的形状(At the edge of the edge detection after get is not necessarily complete, use this function can be a line segment to grow some, and then connect, into the shape of a complete)

说明：  此算法使用于单波段或者多波段图像中的非监督异常检测，性能很好~(The algorithm used in single-band or multi-band image unsupervised anomaly detection, performance very good ~)

此算法为基于k-means聚类的非局部均值快速去噪算法，方法便捷有效，可保持边缘信息(this is a fast edge-preserving denoising method based on non-local means method with k-means clustering)

说明：  图像处理，通过图像融合技术，可以实现将多幅来自同一场景的图像，利用其冗余信息，融合成一幅比原来任何一幅都易于为人们所理解的图像，同时可供人们进行进一步的观察和处理(Through image fusion technology, it can realize the fusion of multiple images from the same scene, and make use of its redundant information to form an image that is easier to understand than any original image)

解决散斑图像的重建问题,主要是使用了相位恢复算法(to study the picture problem)

fpga实现图象滤波，实时的实现现对输入图象的形态学滤波可直接使用。 (fpga image filtering, real-time implementations currently morphological filtering of the input image can be used directly.)

c#下实现图像颜色k均值聚类 将像素点颜色分类并合并(c# implementation of color image k-means clustering to classification of color pixels and merge)

压缩感知CS——采用小波变换进行稀疏表示，高斯随机矩阵为观测矩阵，重构算法为ILRS算法，对256*256的lena图处理，比较原图和IRLS算法在不同采样比例(0.74、0.5、0.3)下的重构效果，并各运行50次，比较算法性能PSNR和每次的运行时间(Compressed sensing CS- using wavelet transform as sparse representation, Gaussian random matrix as the observation matrix and IRLS algorithm as the reconstruction algorithm. Compare lena figure and the reconstruction results using IRLS algorithm at different sampling ratio (0.74,0.50.3),then each runs 50 times, compare the performance of PSNR and each running time)

标准单纯性算法程序,不知能否满足大家需要,()