目录

1 下载原始图片

 2 根据坐标裁剪出人脸子图

编写python脚本下载PubFig人脸公开数据集，每个人的人脸放到一个以人名命名的单独文件夹中，

import os
import sys
import urllib.request as request
from urllib.request import urlretrieve
import socket
import requests
import cv2

'''
下载Pubfig人脸数据集，下载下来之后根据坐标裁剪人脸，
并且每个人的人脸图片放到单独的一个文件夹中
'''

val_urls = "eval_urls.txt"
dev_urls = "dev_urls.txt"

originDir = "./originDir2" #用来保存下载的原始图片
faceDir   = "./faceDir"   #用来保存裁剪的人脸子图


#设定一下无响应时间，
timeout = 3
socket.setdefaulttimeout(timeout)

# 为请求增加一下头
user_agent = 'Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/62.0.3202.62 Safari/537.36'
headers = ('User-Agent', user_agent)
opener = request.build_opener()
opener.addheaders = [headers]
request.install_opener(opener)


'''
    根据url获取文件名字
'''
def getPic(dir, name, nameNum, url, box):
    if not os.path.exists(dir):#每个人的图片放到一个单独的文件夹中，
        os.makedirs(dir, exist_ok=True)
    bad_url = []
    coord = box.split(",")
    print("coord:", coord)
    try:
        fileName = dir + "/" + name + "_" + nameNum + "_" + coord[0] + "_" + coord[1] + "_" + coord[2] + "_" + coord[3] +  ".jpg"
        print("url::", url)
        print("fileName::", fileName)
        request.urlretrieve(url, fileName) 
        #urlretrieve(url, fileName)
    except Exception as e:
        print(Exception, ':', e)
        bad_url.append(url)
  


def downloadPic(txtName):
    with open(txtName) as f:
        lineCount = 0
        nameCount = 0 #用来给图片命名，每个图片的名字为：人名_nameCount
        lines = f.readlines()
        for line in lines:
            #print("line::", line)
            if lineCount >= 2:#txt文件前面两行为数据格式说明，非有效数据，过滤掉。
                #name1, name2, number, url, box, md5 = line.split()  #有的人名有3个单词，这样会报错
                lineList = line.split()
                if 6 == len(lineList):
                    dir     = lineList[0]
                    nameNum = lineList[2]
                    url     = lineList[3]
                    box     = lineList[4]
                if 7 == len(lineList):
                    dir     = lineList[0]
                    nameNum = lineList[3]
                    url     = lineList[4]
                    box     = lineList[5] 
                getPic(originDir + "/" + dir, dir, nameNum, url, box)#目录传进去的是originDir + "/" + dir,
            lineCount = lineCount + 1
            
        

if __name__ == '__main__':
    downloadPic(dev_urls) 
    downloadPic(eval_urls)    
    


     

 2 根据坐标裁剪出人脸子图

同样也是每个人的图片放到一个单独的文件夹中，明明方向是人名加number. 

import os
import sys
import urllib.request as request
from urllib.request import urlretrieve
import socket
import requests
import cv2

'''
根据坐标裁剪出人脸子图
'''

srcDir = './originDir'
dstDir = './faceDir'


def clip(srcDir, dstDir):
    for dirName in os.listdir(srcDir):
        #print("dirname:", dirName)
        for fileName in os.listdir(os.path.join(srcDir, dirName)):                    
            print("filename::",fileName)
            #print("filename::", os.path.join(srcDir, dirName, fileName))
            try:
                originImg = cv2.imread(os.path.join(srcDir, dirName, fileName))
                print("fileName.split('_')[5]:::", fileName.split('_')[5].split('.'))
                personName = fileName.split('_')[0]
                x1 = fileName.split('_')[2]
                y1 = fileName.split('_')[3]
                x2 = fileName.split('_')[4]
                y2 = fileName.split('_')[5].split('.')[0]
                newName = fileName.split('_')[0] + '_' + fileName.split('_')[1] + "." + fileName.split('_')[5].split('.')[1]
                #print("x1::", x1)
                #print("y1::", y1)
                #print("x2::", x2)
                #print("y2::", y2)
                #print("newName::", newName)
                faceImg = originImg[int(y1):int(y2), int(x1):int(x2)]
                faceDir = os.path.join(dstDir, personName)
                print("faceDir::", faceDir)
                if not os.path.exists(faceDir):#每个人的图片放到一个单独的文件夹中，
                    os.makedirs(faceDir, exist_ok=True)
                cv2.imwrite(os.path.join(faceDir, newName), faceImg)
            except Exception as e:
                #print("filename::", os.path.join(srcDir, dirName, fileName))
                print(Exception, ':', e)

if __name__ == '__main__':
    clip(srcDir, dstDir)

参考文献：

Pubfig DataBase 下载Python脚本_CodeCold的博客-CSDN博客_pubfig数据集

python爬取人脸识别图片数据集/py - 云+社区 - 腾讯云

https://www.cnblogs.com/darkknightzh/p/5715305.html