How to scrape high resolution images from google images using BS4 in python

Question

We made a program which accepts an input through a tkinter GUI and goes to google images,and downloads images based on the input.Here is the code:

import requests
import bs4
import random
from PIL import Image
from tkinter import messagebox as msgbox
i=0
import os
from tkinter import *
from tkinter import filedialog
ac=str(random.randint(1,20))
b=str(random.randint(20,38))
y=Tk()
def find_file():
    aaa=filedialog.askdirectory()
    return aaa
def create_folder():
    ad=find_file()
    global ac
    global b
    ads=os.path.join(ad,f"Img{ac}{b}")
    os.mkdir(ads)
    return ads


defe=Entry(bg="white")
defe.grid(row=2,column=2)
adj=Label(text="Enter the name of the photo(s) you want to download :")
adj.grid(row=2,column=1)
ack=Label(text="How many photos you want to download?")
ack.grid(row=3,column=1)
dee=Entry(bg="white")
dee.grid(row=3,column=2)



def download_images():
    defei=defe.get()
    deee=int(dee.get())
    aadgc=[]
    play=True
    if " gif" in defei or ".gif" in defei:
        msgbox.showerror("GIF not supported",".gif format is not supported by this software.Sorry for the inconvenience")
        play=False
    while play:   
        asd=create_folder()
        for start in range(0,400,20):
            bararara=f"https://www.google.co.in/search?q={defei}&source=lnms&tbm=isch&start={start}#imgrc=fTslNdnf0RRRxM"

            a=requests.get(bararara).text
            soup=bs4.BeautifulSoup(a,"lxml")
            ab=soup.find_all("img",{"class":"n3VNCb"},limit=deee)
            aadgc.extend(ab)  
        aa=[abb["src"] for abb in aadgc]
        for source in aa:
            r=random.randint(0,100) 
            ra=random.randint(0,1000) 
            
            raa=asd+"\\"+str(r)+str(ra)+".png"
            try:
                binary=requests.get(source).content
            except requests.exceptions.MissingSchema:
                binary=requests.get("http:"+source).content  
            except:
                binary=requests.get("https:"+source).content       
            with open(raa,"wb") as saaho:
                saaho.write(binary)
                saaho.close()
            global i
            i+=1
            if i==int(deee):
                break 
        
        asd=asd.replace("/","\\")
        os.system(f"explorer \"{asd}\"") 
        break  
aadg=Button(y,bg="red",text="Download!",command=lambda:download_images(),activebackground="dark red",activeforeground="grey")
aadg.grid(row=4,column=1)
y.mainloop()      
aadg=Button(y,bg="red",text="Download!",command=lambda:download_images(),activebackground="dark red",activeforeground="grey")
aadg.grid(row=4,column=1)
y.mainloop() 

But we are getting the thumbnail of the image rather than the image,because of which the software returns only low-resolution photos and doesn't support.gif images.

Also we are not able to find the class that the main image belongs to. Thanks.

Answer 1

With Selenium:

1. Click an image from search results.

2. Wait until the image is visible.

3. image_link = driver.find_element_by_css_selector(".tvh9oe.BIB1wf.eHAdSb>img").get_attribute("src")

You can use the same locator for bs4

Answer 2

To find the original or full resolution image, you have to first get the data-tbnid of an image.

In this case it's: sd7iKvYzujke_M . After you get the ID you just need to use regex to extract the full original image from the page source.

Alternatively you could use a third party solution like SerpApi. It's a paid API with a free trial.

from serpapi import GoogleSearch

params = {
  "api_key": "secret_api_key",
  "engine": "google",
  "q": "inception",
  "tbm": "isch"
}

search = GoogleSearch(params)
results = search.get_dict()

Example JSON output:

"images_results": [
  {
    "position": 1,
    "thumbnail": "https://serpapi.com/searches/60e70bf0e815af01fd163d6a/images/39eac787b1522b4ccc71382ac53fc933e15aa52342a5d06fafca53990897f2f9.jpeg",
    "source": "rottentomatoes.com",
    "title": "Inception (2010) - Rotten Tomatoes",
    "link": "https://www.rottentomatoes.com/m/inception",
    "original": "https://flxt.tmsimg.com/assets/p7825626_p_v10_af.jpg"
  },
  {
    "position": 2,
    "thumbnail": "https://serpapi.com/searches/60e70bf0e815af01fd163d6a/images/39eac787b1522b4ce9c6c6de6d244df2098ae0b7da6856fba545b4376e01d075.jpeg",
    "source": "imdb.com",
    "title": "Inception (2010) - IMDb",
    "link": "https://www.imdb.com/title/tt1375666/",
    "original": "https://m.media-amazon.com/images/M/MV5BMjAxMzY3NjcxNF5BMl5BanBnXkFtZTcwNTI5OTM0Mw@@._V1_.jpg"
  },
  {
    "position": 3,
    "thumbnail": "https://serpapi.com/searches/60e70bf0e815af01fd163d6a/images/39eac787b1522b4ca9d2321fc39291764c3894c9816af84a5af355f2d54f6921.jpeg",
    "source": "screenrant.com",
    "title": "Inception: What Each Character Represents (Confirmed By Christopher Nolan)",
    "link": "https://screenrant.com/inception-movie-christopher-nolan-characters-actors-meaning-confirmed/",
    "original": "https://static2.srcdn.com/wordpress/wp-content/uploads/2020/03/Inception-characters-film-crew.jpg?q=50&fit=crop&w=960&h=500&dpr=1.5"
  },
  ...
]

Check out the documentation for more details.

Disclaimer: I work at SerpApi.

Answer 3

I need to scrape 40,000 search results to build our pharmacy product catalog with a csv file including 40,000+ medication names and barcode, I need to download 5-10 image results per product and name them according to our software indexing. am no coder so I need a software or any help you can give in.