light-center

基于otsu的自适应阀值canny边缘检测(adaptive canny edge detection based on otsu)

1.从图像中识别几何形状的基本方法；基本原理是在于利用点与线的对偶性，将原始图像空间给定的曲线通过曲线表达形式变为参数空间的一个点，这样就把原始图像中给定的检测问题转化为寻找参数空间的峰值问题。 2.基于Gabor变换的人眼检测MATLAB实现。 3.基于Radon的车牌矫正方法。 4.基于模型函数headata产生不同大小的头模型数据。 5.基于函数wdcbm2()设置图像分层阈值压缩参数，实现图像压缩(The basic method of identifying geometry from the image; the basic principle is to use point and line duality, the original image space curve given expression through the curve into a parameter space of a point, so the peak detection problem is given in the original image is transformed into searching the parameter space.)

用区域生长法代码实现对图像的分割，包括阈值的选取(Using region growing method to achieve the image segmentation, including the selection of threshold)

说明：  MATLAB转化大地坐标与球坐标之间的转化，，其中包含了空间直角坐标系与大地坐标系的转化(Matlab transformation of geodetic and spherical coordinates)

剪切波是最近几年发展的图像处理的有利工具，这是几篇不错的论文，希望可以对大家有所帮助。(Shear wave is the last few years the development of image processing tool, this is a few good paper, hope to help everyone.)

医用图像处理程序,处理dicom文件。开源软件,参考dcmtk(dicom process soft)

说明：  基于相邻像素间位异或的图像置乱算法,效果很好，改变图像直方图(Between the adjacent pixels based on the XOR bit of image scrambling algorithm, well, change the image histogram)

读入多组点云数据，全局配准，得到经过配准矫正后的图像数据 文件格式：*.PLY(Read multiple point cloud data, global registration, the registration to be corrected after the image data file formats:*. PLY)

说明：  多种边缘检测算子对比 不需要改动 有sobel prewittts canny log(Multiple edge detection operators)

CS压缩传感的初级教学代码，使用OMP重构，已注释，包括1维信号，2维图像的重构，分别使用dct和小波稀疏，列扫描和分块法进行omp重构(CS compressed sensing primary teaching code using OMP remodeling, already commented, including a 1-dimensional signals, 2-dimensional image reconstruction, respectively, using the DCT and wavelet sparse column scanning and sub-block method omp Reconstruction)