使用Tesseract識別圖片，獲取自如房子價格

mport pytesseract
import requests
import re
from bs4 import BeautifulSoup
from PIL import Image


def get_data():
	# 先請求網頁，獲取基本的數據
    url = 'http://wh.ziroom.com/z/'
    headers = {'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/80.0.3987.87 Safari/537.36'}
    r = requests.get(url, headers).text

    html = BeautifulSoup(r, features='lxml')
    # 查找第一間房子的價格數據
    list_box = html.find('div', attrs={'class': 'Z_list-box'})
    # 獲取所有房子的信息item  這裏只獲取一個
    for item in list_box.find_all('div', attrs={'class': 'item'}):
        break
        
	# 獲取偏移量和圖片的URI
    nums = item.find_all('span', attrs={'class': 'num'})
    positions = []
    for num in nums:
        position = num.get('style').split(':')[-1][:-2]
        uri = re.search(r'\((.*?)\)', num.get('style').split(':')[1]).group(1)
        positions.append(float(position.strip()))

    return positions, uri


def parse_data(positions, uri):
	# 獲取png
    r = requests.get('http:' + uri).content
    with open('img.png', 'wb') as f:
        f.write(r)
	# 讀取
    img = Image.open('img.png')
    img.show()

    # 初始化
    pytesseract.pytesseract.tesseract_cmd = r'D:\Tesseract-OCR\tesseract.exe'
    # 配置
    config = r'--tessdata-dir "D:\Tesseract-OCR\tessdata"'

    # 識別
    nums = pytesseract.image_to_string(img, lang='eng', config=config).strip()
    # 識別的時候會多一些空格，去掉 
    n = [num for num in nums if num != ' ']
    print(n)

    # 根據偏移量獲取價格的數字
    positions = [int(abs(position / 21.4)) for position in positions]
    price = ''
    for p in positions:
        price += n[p]

    print(price)

if __name__ == '__main__':
    p, u = get_data()
    parse_data(p, u)

準確率很高，試了很多次結果都是對的上的，運行效果：

不想寫怎麼分析，人！懶！該！打！具體分析可參見下鏈接，前輩的分析，很全~~

https://blog.csdn.net/weixin_45081575/article/