#!/bin/python
#coding:utf8
import os
import exifread
import re
import sys
import requests
import json__author__ = 'DaDaLuLa'#************************************************************************
#代码功能：#
#   1.读取所有图片文件的exif信息                                            #
#   2.提取图片中的经纬度，将度、分、秒转换为小数形式                           #
#   3.利用百度地图API接口将经纬度转换成地址形式                               #
#************************************************************************#遍历文件夹及子文件夹中的所有图片,逐个文件读取exif信息
def get_pic_GPS(pic_dir):items = os.listdir(pic_dir)for item in items:path = os.path.join(pic_dir, item)if os.path.isdir(path):get_pic_GPS(path)else:imageread(path)#将经纬度转换为小数形式
def convert_to_decimal(*gps):#度if '/' in gps[0]:deg = gps[0].split('/')if deg[0] == '0' or deg[1] == '0':gps_d = 0else:gps_d = float(deg[0]) / float(deg[1])else:gps_d = float(gps[0])#分if '/' in gps[1]:minu = gps[1].split('/')if minu[0] == '0' or minu[1] == '0':gps_m = 0else:gps_m = (float(minu[0]) / float(minu[1])) / 60else:gps_m = float(gps[1]) / 60#秒if '/' in gps[2]:sec = gps[2].split('/')if sec[0] == '0' or sec[1] == '0':gps_s = 0else:gps_s = (float(sec[0]) / float(sec[1])) / 3600else:gps_s = float(gps[2]) / 3600decimal_gps = gps_d + gps_m + gps_s#如果是南半球或是西半球if gps[3] == 'W' or gps[3] == 'S' or gps[3] == "83" or gps[3] == "87":return str(decimal_gps * -1)else:return str(decimal_gps)#读取图片的经纬度和拍摄时间
def imageread(path):f = open(path,'rb')GPS = {}Data = ""try:tags = exifread.process_file(f)except:return'''for tag in tags:print(tag,":",tags[tag])'''#南北半球标识if 'GPS GPSLatitudeRef' in tags:GPS['GPSLatitudeRef'] = str(tags['GPS GPSLatitudeRef'])else:GPS['GPSLatitudeRef'] = 'N' #缺省设置为北半球#东西半球标识if 'GPS GPSLongitudeRef'in tags:GPS['GPSLongitudeRef'] = str(tags['GPS GPSLongitudeRef'])else:GPS['GPSLongitudeRef'] = 'E' #缺省设置为东半球#海拔高度标识if 'GPS GPSAltitudeRef' in tags:GPS['GPSAltitudeRef'] =  str(tags['GPS GPSAltitudeRef'])#获取纬度if 'GPS GPSLatitude' in tags:lat = str(tags['GPS GPSLatitude'])#处理无效值if lat == '[0, 0, 0]' or lat == '[0/0, 0/0, 0/0]':returndeg, minu, sec = [x.replace(' ', '') for x in lat[1:-1].split(',')]#将纬度转换为小数形式GPS['GPSLatitude'] = convert_to_decimal(deg, minu, sec,GPS['GPSLatitudeRef'])#获取经度if 'GPS GPSLongitude' in tags:lng = str(tags['GPS GPSLongitude'])#处理无效值if lng == '[0, 0, 0]' or lng == '[0/0, 0/0, 0/0]':returndeg, minu, sec = [x.replace(' ', '') for x in lng[1:-1].split(',')]#将经度转换为小数形式GPS['GPSLongitude'] = convert_to_decimal(deg, minu, sec,GPS['GPSLongitudeRef'])#对特殊的经纬度格式进行处理#获取海拔高度if 'GPS GPSAltitude' in tags:GPS['GPSAltitude'] = str(tags["GPS GPSAltitude"])#获取图片拍摄时间if 'Image DateTime' in tags:GPS["DateTime"] = str(tags["Image DateTime"])elif "EXIF DateTimeOriginal" in tags:GPS["DateTime"] = str(tags["EXIF DateTimeOriginal"])if 'GPSLatitude' in GPS:#将经纬度转换为地址convert_gps_to_address(GPS)#利用百度全球逆地理编码服务（Geocoder）Web API接口服务将经纬转换为位置信息
def convert_gps_to_address(GPS):secret_key = 'xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx'#百度密钥lat, lng = GPS['GPSLatitude'], GPS['GPSLongitude']#注意coordtype为wgs84ll(GPS经纬度),否则定位会出现偏差baidu_map_api = "http://api.map.baidu.com/geocoder/v2/?coordtype=wgs84ll&location={0},{1}&output=json&pois=0&ak={2}".format(lat,lng,secret_key)content = requests.get(baidu_map_api).textgps_address = json.loads(content)#结构化的地址formatted_address = gps_address["result"]["formatted_address"]#国家（若需访问境外POI，需申请逆地理编码境外POI服务权限）country = gps_address["result"]["addressComponent"]["country"]#省province = gps_address["result"]["addressComponent"]["province"]#城市city = gps_address["result"]["addressComponent"]["city"]#区district = gps_address["result"]["addressComponent"]["district"]#语义化地址描述sematic_description = gps_address["result"]["sematic_description"]#将转换后的信息写入文件with open("gps_address.csv","a+") as csv:csv.write(GPS["DateTime"] + "|" + formatted_address + "|" + country + "|" + province + "|" + city + "|" + district + "|" + sematic_description + "n")if __name__ == "__main__":get_pic_GPS("./photo/")

#代码在notebook中执行
%matplotlib inline
import pandas as pd
import matplotlib.pyplot as plt
import numpy as np
plt.rcParams['font.sans-serif'] = ['Simhei']
df = pd.read_csv('gps_address.csv',sep='|')
fig = plt.figure(figsize=(15,20))
data = df["province"].value_counts()
num_of_province = data.values
plt.barh(range(len(num_of_province),0,-1),num_of_province,height=0.7,color='steelblue',alpha=0.8)
plt.title("各省分布情况",fontsize=20)
plt.xlabel("数量",fontsize=15)
plt.ylim(0,len(num_of_province)+1)
plt.yticks(range(len(num_of_province),0,-1),data.index,fontsize=15)
for x,y in enumerate(np.sort(num_of_province)):plt.text(y + 0.5,x + 0.9,'%s' % y,fontsize=18 )
plt.show()

fig = plt.figure(figsize=(15,40))
data = df["city"].value_counts()
num_of_city = data.values
plt.barh(range(len(num_of_city),0,-1),num_of_city,height=0.8,color='steelblue',alpha=0.8)
plt.title("城市分布情况",fontsize=20)
plt.xlabel("数量",fontsize=15)
plt.yticks(range(len(num_of_city),0,-1),data.index,fontsize=15)
plt.ylim(0,len(num_of_city)+2)
for x,y in enumerate(np.sort(num_of_city)):plt.text(y +0.2,x + 0.8,'%s' % y,fontsize=18 )
plt.show()

from pyecharts import Map, Geo
import re
provinces = list(df["province"].value_counts().index)
provinces = [re.sub("壮族自治区|自治区|省|市","",x) for x in provinces]
print(provinces)pro_values = list(df["province"].value_counts().values)
city = list(df["province"].value_counts().index)
city_values = list(df["province"].value_counts().values)geo = Geo("各省分布情况", "", title_color="#fff",title_pos="center", width=1000,height=600, background_color='#404a59')
geo.add("", provinces, pro_values, visual_range=[0, 200], maptype='china',visual_text_color="#fff",symbol_size=10, is_visualmap=True)geo

city = list(df["city"].value_counts().index)
city= [re.sub("白族自治州|市","",x) for x in city]
city_values = list(df["city"].value_counts().values)
geo = Geo("城市分布情况", "", title_color="#fff",title_pos="center", width=1000,height=600, background_color='#404a59')
geo.add("", city, city_values, visual_range=[0, 200], maptype='china',visual_text_color="#fff", symbol_size=10, is_visualmap=True)
geo