Search Engines Image Scraper

You do want to find high quality image with custom setting? here is what you need with Search Engines Image Scraper You can easily find and download image with custom size , color and ratio. Let's start and Start project if you want help to improve It's My honor to check Your pull request.

Table of Content

1. requirement and installation
2. Usage
   • Cli
   • Pythn module
3. Defindes
   • image ratio
   • image size
   • image color
   • table of paramters
   • Why DuckDuckGo?
4. helpful link

---

requirement:

1. python 3+ (it can use also for python 2 but need to change)
2. python packges: selenium , numpy , cv2

install requirement (freeze version):

• install with pip install -r requirement.txt' or pip3 install -r requirement.txt

---

Usage:

1. CLI

open Your shell in repo dir and enter (python3 also for some cases)

CLI version just accept two paramanter

python image_scraper.py --qeury string --amount int

for sample here I am download amazon logo

python image_scraper.py --qeury "amazon logo" --amount 10

for get help wirte:

python image_scraper.py -h

2. python module

1. Import it as normal module

from image_scraper import Scraper_Config , Image_Scraper

2. Create config (also you can use default config)

config = Scraper_Config(SEARCH_QUERY = 'duckduckgo',
                 NUMBER_OF_PICTURES = 5,
                 IMAGE_RATIO = (16, 9),
                 IMAGE_SIZE = (200 , 100),
                 CHECK_RATIO_AND_RESIZE = False,
                 EPSILON_RATIO_ERROR = 0.5 ,
                 MAKE_GRAY = False,
                 SEARCH_ENGINE_TYPE='duckduckgo',
                 MINIMUM_ELEMENT_ERROR = 0.2,
                 SLEEP_TIME = 0,
                 SCROLL_PAUSE_TIME = 2)

3. Create Image_Scraper isntace and call scrap function

scraper = Image_Scraper(config)
scraper.scrap(verbose=True)

---

Defindes

aspect ratios

what is aspect ratio and why is important?

Maybe you known aspect ratio in images but why is important to us?

Let's check an example, our task is find great mohammad ali picture from google or duckduckgo. in this system you can easily define image ratio and which size if ratio with some gap is neer to ratio accept it else next one. look at paramter table for more details.

parameter table for better config

Parameter	Type	Usage	Example
SEARCH_ENGINE_TYPE	one of 'google', 'duckduckgo'	Chose use which search engine	duckduckgo
NUMBER_OF_PICTURES	Int > 5	how much will be amount of images	5
CHECK_RATIO_AND_RESIZE	Boolean	If True then check ratio and resize each image	True
IMAGE_RATIO	tuple each number	If CHECK_RATIO_AND_RESIZE true then need to spesify it what is ratio	(16, 9)
EPSILON_RATIO_ERROR	float greater than 0.1	If CHECK_RATIO_AND_RESIZE true then need to check until how much bigger ratio can resize
GOOGLE_URL	String	Url for open in browser	as is in default
DUCKDUCKGO_URL	String	Url for open in browser	as is in default
SCROLL_PAUSE_TIME	Int -> second	Wait between every scroll for load in browser	2 (seconds)
PATH_OF_THIS_FILE	String -> path	image_scraper.py file path	as is in default
IMAGE_BASE_DIR	String -> path	where image save	as is in default
GOOGLE_DRIVER_PATH	Int -> second	If need to stop for download every image then put greater than 1	10 (seconds)
GOOGLE_CHROME_DRIVER_PATH	String -> path	Path of google chrome driver	as is in default
IMAGE_RATIO_PERSENT	Float	convert ratio to float number	1.7 as mohammd ali picture
MAKE_GRAY	Boolean	If CHECK_RATIO_AND_RESIZE true then make images gray	False
MINIMUM_ELEMENT_ERROR	Float	if you need 100 image then at least check exist of 120 image this var is gap between 100 to 120. also when check CHECK_RATIO_AND_RESIZE need up to 2 time more elemnt to check and download needed image with custom ratio	0.2
MINIMUM_EMELMET	Int	calculate number of minimum elemnt must be exist	as is in default
HTML_EMELMT_CLASS_NAME	String -> html tag	Html tag in page for check and donwload image	10 (seconds)
USE_MULTI_THREADING	Boolean	Use multi threading or not	True
RANDOM_SEED	Int	What is Your random seed for stop between threads or after done processes	420

Why DuckDuckGo?

duckduckgo say "Privacy, simplified ..." why here I am not talking about Privacy here we need Scrap web. Google Check ip, Behivaer in Browser and almost all thing with Privacy option you can even scrap it with TOR network

helpful link

some issue I got maybe helpful for you:

• FileNotFoundError: [Errno 2] No such file or directory: 'geckodriver': 'geckodriver' link 1 stackoverflow - link 2 askubuntu

• install last Chrome on ubuntu

• install Chrome web driver ubuntu --

todo list

• add more example
• add google search engine
• add more browser driver
• add batch image downloader
• add image instant augment
• add multi threading support
• add just resize without checking ratio
• add colab support and colab file exmaple