kecoefficient

很实用的群优化智能算法，计算粒子群算法的源代码，所需优化的目标函数命名为fitness即可。(a partical swarm algorithm)

sar图像去噪的几种新的方法，通过反复训练模板能有较高的识别率，快速扩展随机生成树算法。( Several new methods sar image denoising, Through repeated training XyjPRvRlate have higher recognition rate, Rapid expansion of random spanning tree algorithm.)

去斜处理（宽带的匹配滤波）Stretch方法是针对线性调频信号而提出的， 其方法是将输入信号与参考信号 （经适当延迟的本振信号， 延迟量通常由窄带信号测距结果估计出） 混频 ，则每一个散射点就对应一个混频后的单频分量，对混频输出的信号进行 DFT 处理 ，即可获得目标的距离像，对参考信号的要求是应具有与输入信号相同的调频斜率。(dechirp matched filter Stretch method is proposed for linear frequency modulation signal. It mixes the input signal with the reference signal (local oscillator signal with appropriate delay, the delay is usually estimated from the result of narrowband signal ranging). Then each scattering point corresponds to a single frequency component after mixing, and DFT is performed on the mixed-frequency output signal. The range profile of the target can be obtained by processing, and the reference signal is required to have the same frequency modulation slope as the input signal.)

我们的目的是将一个分数拆分成若干个分子是1分母是正整数的分数之和的形式。例如2/3可以拆分成1/2+1/6的形式。这里不考虑顺序，比如1/6+1/2和1/2+1/6相同。 输入p, q, a, n, 其中p/q为分数，对这个分数进行拆分，要求拆分的分数个数不能多于n个，所有分母的乘积不能大于a。求有多少种拆分方法? 输入：最多200组数据。在每行中都有四个数分别表示p, q, a, n，之间以空格隔开，并且满足p,q<=800，a<=12000，n<=7，最后以四个0做为输入的结束。 输入示例: 2 3 120 3 2 3 300 3 0 0 0 0 输出示例: 4 7 (Our aim is to split a fraction into a number of elements is a denominator is a positive integer fraction of the sum of form. For example, 2/3 can be split into 1/2+1/6 form. Not considered here, the order, such as 1/6+1/2 and 1/2+1/6 of the same. Input p, q, a, n, where p/q as a fraction of this split scores require no more than the number of split points n-all the product of the denominator can not be larger than a. How many kinds of demand split method? Input: Up to 200 sets of data. In each line has four numbers indicate the p, q, a, n, separated by a space between, and meet the p, q < = 800, a < = 12000, n < = 7, the end, four 0 as the input end. Input Example: 2,312,032,330,030,000 sample output: 47)

说明：  摄像机标定算法的MATLAB语言实现，通过MATLAB实现对摄像机的标定(Camera calibration algorithm of the MATLAB language, through the MATLAB camera calibration to achieve)

这是一个关于卷积编码和Viterbi译码的Matlab程序，现在把它们放在这里，希望对需要的人有些帮助。(This is a convolution coding and Viterbi decoding of Matlab procedures, now put them in here, I hope those who need some help.)

通过输入相关参数，实现三阶累积量的求法。从而避免复杂的计算过程。(By input parameters, the third order cumulant method. To avoid complex calculation process. )

nvec_power: power supply driver for a NVIDIA compliant embedded controller.

UWB接收性能仿真 包括高斯信道 模板信号 误码率(UWB receiver performance simulation include the Gaussian channel template signal bit error rate)

这段程序是关于图像分析中的各种颜色处理的，感兴趣的同学可以看看(This program is all about the color of the image analysis processing, interested students can look)