psotb-beta-0.3

基于kalman滤波的目标跟踪代码，用matlab实现了对目标的跟踪。(Kalman filter based target tracking code, matlab achieve the target tracking.)

四步搜索算法对三步算法进行了适当的改进，在一定程度上弥补了TSS算法在估计细小运动时的不足，并加入了半途中止的策略，进一步提高了搜索精度 (Four-step search algorithm for three-step algorithm to improve the appropriate, to a certain extent make up for the TSS algorithm in estimating the inadequacy of a small movement, and joined the mid-suspension strategy to further improve the search accuracy)

ch3=softt(h3,thr1 clear clc I=imread( lena512.bmp ) imshow(I) I=double(I) II=I+30*randn(size(I)) figure imshow(II,[]) [a1,h1,v1,d1]=dwt2(II, sym8 ) [a2,h2,v2,d2]=dwt2(a1, sym8 ) [a3,h3,v3,d3]=dwt2(a2, sym8 ) sigma=median(abs(d1(:)))/0.6745 x=3*sigma thr1=x*2^(-1) thr2=x*2^(-0.5) thr3=x 小波图像去噪( ch3=softt(h3,thr1 clear clc I=imread( lena512.bmp ) imshow(I) I=double(I) II=I+30*randn(size(I)) figure imshow(II,[]) [a1,h1,v1,d1]=dwt2(II, sym8 ) [a2,h2,v2,d2]=dwt2(a1, sym8 ) [a3,h3,v3,d3]=dwt2(a2, sym8 ) sigma=median(abs(d1(:)))/0.6745 x=3*sigma thr1=x*2^(-1) thr2=x*2^(-0.5) thr3=x )

Unscented Particle Filter源程序(Unscented Particle Filter PROGRAM)

matlab编写的多示例多标记分类程序，采用miSVM算法，svm做分类器，其中附有各个调用函数，和作为例子的数据和主程序，可以运行，可以多多示例多标记样本进行分类。(matlab prepared multi-instance multi-label classification procedures, using miSVM algorithm, svm do classifier, which function with each call, and as an example of the data and the main program, you can run, you can sample a lot of examples of multi-label classification.)

遗传算法GA—模拟自然界遗传机制和生物进化论而成的一种并行随机搜索最优化方法。（具有“生存+检测”的迭代过程的搜索算法）基于自然界“优胜劣汰，适者生存”的生物进化原理引入优化参数形成的编码串联群体中，按照所选择的适应度函数并通过遗传中的选择、交叉和变异对个体进行筛选，使适应度值好的个体被保留，适应度差的个体被淘汰，新的群体既继承了上一代的信息，又优于上一代。反复循环，直至满足条件。(GA-A parallel random search optimization method that simulates natural genetic mechanisms and biological evolution. (Search algorithm with "survival + detection" iterative process) Based on the biological evolution principle of "survival of the fittest, survival of the fittest" in the natural world, the coding series formed by optimized parameters is introduced. Individuals are screened according to the selected fitness function and through genetic selection, crossover, and variation, so that individuals with good fitness values are retained, individuals with poor adaptability are eliminated, and new groups inherit the information of the previous generation., Better than the last generation. Repeat until conditions are met.)

Bidirectional associative memory does heteroassociative processing. Here BAM algorithm is implemented.

support vector classification

线性调频信号的时域，频域表示，包括坐标转换设置。对雷达初学者有指导意义！(Linear FM signals in time domain, frequency domain representation, including coordinate conversion settings. Radar beginners guide!)

invetre 3phase by jahed