三维点云中DBSCAN的使用

# -*- coding: utf-8 -*-
import numpy as np
import matplotlib.pyplot as plt
from mpl_toolkits.mplot3d import Axes3D

size = 30

##计算欧式距离
def distEuclid(x,y):
    return np.sqrt(np.sum((x-y)**2))

##随机产生n个dim维度的数据 (这里为了展示结果 dim取2或者3)
def genDataset(n,dim):
    data = []
    while len(data)<n:
        p = np.around(np.random.rand(dim)*size,decimals=2)
        data.append(p)
    return data

##判断两点是否在范围内
def isNeighbor(x,y,eps):
    return distEuclid(x,y)<=eps

##获取某一点邻域内的点
def getSeedPos(pos,data,eps):
    seed = []
    for p in range(len(data)):
        if isNeighbor(data[p],data[pos],eps):
            seed.append(p)
    return seed

##获取核心点列表
def getCorePointsPos(data,eps,minpts):
    cpoints=[]
    for pos in range(len(data)):
        if len(getSeedPos(pos,data,eps))>=minpts:
            cpoints.append(pos)
    return cpoints

##分类
def getCluster(data,eps,minpts):
    corePos = getCorePointsPos(data,eps,minpts)
    unvisited =list(range(len(data)))
    cluster = {}
    num = 0
    
    for pos in corePos:
        if pos not in unvisited:
            continue
        clusterpoint = []
        clusterpoint.append(pos)
        seedlist = getSeedPos(pos,data,eps)
        unvisited.remove(pos)
        while seedlist:
            p = seedlist.pop(0)
            if p not in unvisited:
                continue
            unvisited.remove(p)
            clusterpoint.append(p)
            if p in corePos:
                seedlist.extend(getSeedPos(p,data,eps))
        cluster[num] = clusterpoint    
        num+=1
    cluster["noisy"]=unvisited
    return cluster

##展示结果  各类簇使用不同的颜色  中心点使用X表示
def Show(data,cluster):
    num,dim = data.shape
    color = ['r','g','c','y','m','b','pink','maroon','tomato','peru','lawngreen','gold','aqua','dodgerblue']
    ##二维图
    if dim==2:
        for i in cluster:
            pos = cluster[i]
            if i=="noisy":
                for p in pos:
                    plt.plot(data[p,0],data[p,1],'o',c='k')
            else:
                for p in pos:
                    plt.plot(data[p,0],data[p,1],'o',c=color[i])            
    ##三维图
    elif dim==3:
        ax = plt.subplot(111,projection ='3d')
        for i in cluster:
            pos = cluster[i]
            if i=="noisy":
                for p in pos:
                    ax.scatter(data[p,0],data[p,1],data[p,2],c='black')
            else:
                for p in pos:
                    ax.scatter(data[p,0],data[p,1],data[p,2],c=color[i])
        
    plt.show()
        
        

data = np.array(genDataset(80,3))
print(type(data))
print(data.shape)
cl = getCluster(data,6,4)
Show(data,cl)

标签：dim,DBSCAN,##,云中,三维,pos,cluster,unvisited,data
来源： https://blog.csdn.net/minhuaQAQ/article/details/118345703