Calib_main

iNEDI is an acronym for "improved New Edge-Directed Interpolation". This function implements an improvement of the original NEDI technique described on: X. Li et. al., "new edge-directed interpolation". IEEE Trans. on Image Processing, Vol.10, No.10, October 2001, pp.1521-1527 (iNEDI is an acronym for "improved New Edge- Directed Interpolation. " This function imple ments an improvement of the original GaAs techn ique described on : X. Li et. al. , "new edge-directed interpolation." IEEE Tran s. on Image Processing, Vol.10, No.10, 373,890 bytes, pp.1521- 1527)

对于抽样定理的验证，十分好，二十字不太好的(For the sampling theorem verification, very good, not good twenty words)

bp神经网络的设计 希望大家相互探讨 相互学习 (bp neural network design hope we mutually explore and learn from each other)

MATLAB实现信号的盲源分离~fastICA方法~源信号与混合信号个数均为2~(MATLAB for signal blind source separation method ~ ~ fastICA number of source signals and mixed signals are 2 ~)

对图像分割，初始轮廓为椭圆。变化的过程中始终以椭圆为约束(For image segmentation, the initial contour is an ellipse.)

这是一个用c和matlab编写的程序，用于实现k-means算法(This is a matlab with c and the preparation of the procedures used to implement k-means algorithm)

粒子群算法优化广义回归神经网络的代码，具有很好的借鉴意义。(Particle swarm optimization (PSO) algorithm optimizes the code of generalized regression neural network, which has a good reference value.)

对HARQ系统的吞吐量分析，模式识别中的bayes判别分析算法，粒子图像分割及匹配均为自行编制的子例程，给出接收信号眼图及系统仿真误码率，BP神经网络用于函数拟合与模式识别，用于信号特征提取、信号消噪，合成孔径雷达（SAR)目标成像仿真，有CDF三角函数曲线/三维曲线图。( HARQ throughput analysis of the system, Pattern Recognition bayes discriminant analysis algorithm, Particle image segmentation and matching subroutines themselves are prepared, The received signal is given eye and BER simulation systems, BP neural network function fitting and pattern recognition, For feature extraction, signal de-noising, Synthetic Aperture Radar (SAR) imaging simulation target, There CDF trigonometric curve/3D graphs.)

一种模糊分类算法—基于RBF的扩展RCE算法，空间复杂度低，易于在实时系统中应用(A fuzzy classification algorithm- an extension based on RBF RCE algorithm, space complexity is low, easy to apply in real-time systems)

两个程序：2层BP网络以及3层BP网络实现的函数逼近。(Two procedures: 2-layer BP network, as well as 3-layer BP network s function approximation.)