Crawl Original Google Images & Youtube Videos

---

This repo contains code to crawl images and videos:

• ORIGINAL images from Google Search
• ORIGINAL videos from Youtube

Requirements

1. ChromeDriver

   • Check your current Google Chrome Version
   • Download ChromeDriver corresponding to your Chrome Version at ChromeDriver, unzip it.

   For example, I'm using Chrome Version 95.0.4638.69, Linux, so I downloaded chromedriver_linux64.zip

2. Enviroments conda env create -f environment.yml

Crawl Images from Google Image Search

Download original (not thumbnails) from Google Images Search with multi-threading :D

1. Get URLs by keywords

   	python crawl_url.py
   

2. Download imgs from URLs

   	python crawl_data.py
   

Crawl Videos from Youtube

1. Get URLs by keywords

   python crawl_youtube_link.py
   

2. Download videos from URLs

   python crawl_videos.py
   python crawl_videos.py --metadata --thumbnail # thumbnail and metadata only
   

To-do

• Init
• Multithreading
• Requiremets
• Write Guideline
• Add parser to save_dirs, chromedriver, etc.