搜索资源列表
LDFV
- VLAD VLAD可以理解为是BOF和fisher vector的折中 BOF是把特征点做kmeans聚类,然后用离特征点最近的一个聚类中心去代替该特征点,损失较多信息; Fisher vector是对特征点用GMM建模,GMM实际上也是一种聚类,只不过它是考虑了特征点到每个聚类中心的距离,也就是用所有聚类中心的线性组合去表示该特征点,在GMM建模的过程中也有损失信息; VLAD像BOF那样,只考虑离特征点最近的聚类中心,VLAD保存了每个特征点到离它最近的聚类中心的距离;
test
- 对于人类运行轨迹数据,将轨迹的停留点找到并进行聚类分析-For the trajectory of human data, will stay point trajectory find and cluster analysis
hangkie
- 计算一维光子晶体的透射特性和反射特性,复化三点Gauss-lengend公式求pi,可实现对二维数据的聚类。- Calculated transmission characteristics and reflection characteristics of the one-dimensional photonic crystals, Complex of three-point Gauss-lengend the Formula pi, Can realize the two-dimensio
muimeng_v23
- 二维声子晶体FDTD方法计算禁带宽度的例子,用MATLAB实现动态聚类或迭代自组织数据分析,表示出两帧图像间各个像素点的相对情况。- Dimensional phononic crystals FDTD method calculation examples band gap, Using MATLAB dynamic clustering or iterative self-organizing data analysis, Between two images showing the rel
M8激光雷达动目标跟踪仿真
- 模拟八线激光雷达产生点云数据,实现目标聚类,并对聚类的目标进行跟踪。(Simulated eight line lidar generates point cloud data, achieves target clustering, and tracks the clustering targets.)
PSO_Kmeans
- K-means聚类算法,基于PSO改的聚类算法,对初始点的选择进行优化(K-means clustering algorithm, based on PSO modified clustering algorithm, optimized the selection of the initial point.)
k-means程序
- 介绍了k-means 均值聚类,能很好的将离散的点,聚类成几个指定的聚合点。(The K-means mean clustering is introduced, and the discrete points can be well clustered into several designated aggregation points.)
dbscan
- 有代表性的基于密度的聚类算法 ,将簇定义为密度相连的点的最大集合,能够把具有足够高密度的区域划分为簇,并可在噪声的空间数据库中发现任意形状的聚类。(DBSCAN(Density-Based Spatial Clustering of Applications with Noise))