atom1

边缘是图像最基本的特征，是图像分割的第一步。经典的边缘检测方法如：Roberts，Sobel，Prewitt，Kirsch，Laplace等方法，基本都是对原始图像中象素的小邻域构造边缘检测算子，进行一阶微分或二阶微分运算，求得梯度最大值或二阶导数的过零点，最后选取适当的阀值提取边界。由于这些算法涉及梯度的运算，因此均存在对噪声敏感、计算量大等缺点。在实践中，发现SUSAN算法只基于对周边象素的灰度比较，完全不涉及梯度的运算，因此其抗噪声能力很强，运算量也比较小。并将SUSAN算法用于多类图像的边缘检测中，实验证明该算法非常适合含噪图像的边缘检测。 2 SUSAN边缘检测简介 (err)

matlab GUI界面B样条曲线绘制，内附说明文档,可以移动控制点(B spline curve drawing program with matlab GUI interface )

This is a matrix multiplication as an example of OpenMP parallel programming.

Adaptive background mixture models for real-time tracking

非常不错的QPSK在matlab中的仿真教程，文档中包含相关的所有代码！(Very good QPSK Simulation in matlab tutorial document contains all the code!)

说明：  利用Matlab编程验证用EM算法估计的高斯混合模型的相关参数的性能。(Validate the use of Matlab programming estimated using EM algorithm for Gaussian mixture model parameters related to the performance.)

结合carsim-simulink，设计路径跟踪算法。(Combined with carsim-simulink, the path tracking algorithm is designed.)

说明：  该程序包是美国某大学的团队联合编写的潮流计算程序，包含高司法、牛拉法、PQ法等，对电力系统的潮流计算描述的非常完善。(The package is a university in the United States jointly prepared the team flow calculation program, including high-judicial, cattle Rafah, PQ law, the power flow calculation is described very well.)

数据联合检测与信道估计源码，此为博士论文相关程序，对信道及信号检测有详尽的描述和仿真(Joint data detection and channel estimation source, this doctoral thesis program, a detailed description of the channel and the signal detection and simulation)

Kalman滤波处理运动小球圆心位置，实时测量小球的速度、加速度(Kalman filtering processing ball movement right position )