原文为JS版本，见JS版本
这里将其转成python版本，如下：
from geog import distance
import pandas as pd
from matplotlib import pyplot as plt
import numpy as np
class Douglas_Peuker:
"Douglas–Peucker algorithm压缩轨迹"
def __init__(self, traj, dMax=5):
print('原始轨迹长度：', len(traj))
plt.figure() # 画图
plt.subplot(1, 2, 1) # 画图
plt.plot(np.array(traj)[:,0], np.array(traj)[:,1], 'r-')
plt.title('init_traj')
traj = [[i, x[0], x[1]] for i, x in enumerate(traj)]
douglasPecker_traj = self.douglasPecker(traj, dMax)
print('压缩后轨迹长度：', len(douglasPecker_traj))
plt.subplot(1, 2, 2) # 画图
plt.plot(np.array(douglasPecker_traj)[:, 0], np.array(douglasPecker_traj)[:, 1], 'b-')
plt.title('douglasPecker_traj')
plt.show()
plt.close()
def douglasPecker(self, coordinate, dMax):
"""
:param coordinate: 原始轨迹
:param dMax: 允许最大距离误差
:return: douglasResult 抽稀后的轨迹
"""
result = self.compressLine(coordinate, [], 0, len(coordinate)-1, dMax)
result.append(coordinate[0])
result.append(coordinate[-1])
result = pd.DataFrame(result, columns=['myIndex', 'lat', 'lng'])
result.sort_values(by=['myIndex'], inplace=True)
result.reset_index(drop=True, inplace=True)
return result.loc[:, ['lat', 'lng']].values.tolist()
def compressLine(self, coordinate, result, start, end, dMax):
"递归方式压缩轨迹"
if start<end:
maxDist = 0
currentIndex = 0
startPoint = coordinate[start][1:]
endPoint = coordinate[end][1:]
for i in range(start+1, end):
currentDist = self.disToSegment(startPoint, endPoint, coordinate[i][1:])
if currentDist>maxDist:
maxDist = currentDist
currentIndex = i
if maxDist>=dMax:
# 将当前点加入到过滤数组中
result.append(coordinate[currentIndex])
# 将原来的线段以当前点为中心拆成两段，分别进行递归处理
self.compressLine(coordinate, result, start, currentIndex, dMax)
self.compressLine(coordinate, result, currentIndex, end, dMax)
return result
def disToSegment(self, start, end, center):
"计算垂距，用海伦公式计算面积"
a = distance(start, end)
b = distance(start, center)
c = distance(end, center)
p = (a+b+c) / 2
s = np.sqrt(abs(p*(p-a)*(p-b)*(p-c)))
return s*2/a
if __name__=='__main__':
"traj是一个二维数组格式的经纬度列表,lng在前，lat在后"
Douglas_Peuker(traj, dMax=5)
结果：
原始轨迹长度： 167
压缩后轨迹长度： 129