Author
Last Commit
Oct. 23, 2018
Created
Mar. 28, 2015

Contents

## Summary

This is a command line python program to search keywords/key-phrases on Google Images and optionally download images to your computer. You can also invoke this script from another python file.

This is a small and ready-to-run program. No dependencies are required to be installed if you would only want to download up to 100 images per keyword. If you would want more than 100 images per keyword, then you would need to install Selenium library along with chromedriver. Detailed instructions in the troubleshooting section.

## Compatibility

This program is compatible with both the versions of python - 2.x and 3.x (recommended). It is a download-and-run program with no changes to the file. You will just have to specify parameters through the command line.

## Installation

You can use one of the below methods to download and use this repository.

Using pip

$pip install google_images_download Manually using CLI $ git clone https://github.com/hardikvasa/google-images-download.git
$cd google-images-download && sudo python setup.py install Manually using UI Go to the repo on github ==> Click on 'Clone or Download' ==> Click on 'Download ZIP' and save it on your local disk. ## Usage - Using Command Line Interface If installed via pip or using CLI, use the following command: $ googleimagesdownload [Arguments...]

$python3 google_images_download.py [Arguments...] OR$ python google_images_download.py [Arguments...]

## Usage - From another python file

If you would want to use this library from another python file, you could use it as shown below:

from google_images_download import google_images_download

absolute_image_paths = response.download({<Arguments...>})

## Arguments

Argument Short hand Description
config_file cf

You can pass the arguments inside a config file. This is an alternative to passing arguments on the command line directly.

Please refer to the config file format below

• If 'config_file' argument is present, the program will use the config file and command line arguments will be discarded
• Config file can only be in JSON format
• Please refrain from passing invalid arguments from config file. Refer to the below arguments list
keywords k

Denotes the keywords/key phrases you want to search for. For more than one keywords, wrap it in single quotes.

Tips:

• If you simply type the keyword, Google will best try to match it
• If you want to search for exact phrase, you can wrap the keywords in double quotes ("")
• If you want to search to contain either of the words provided, use OR between the words.
• If you want to explicitly not want a specific word use a minus sign before the word (-)
keywords_from_file kf

Denotes the file name from where you would want to import the keywords.

Add one keyword per line. Blank/Empty lines are truncated automatically.

Only file types '.txt' or '.csv' are allowed.

prefix_keywords pk

Denotes additional words added before main keyword while making the search query.

The final search query would be: <prefix keyword> <keyword>

So, for example, if the keyword is 'car' and prefix_keyword is 'red,yellow,blue', it will search and download images for 'red car', 'yellow car' and 'blue car' individually

suffix_keywords sk

Denotes additional words added after main keyword while making the search query.

The final search query would be: <keyword> <suffix keyword>

So, for example, if the keyword is 'car' and suffix_keyword is 'red,yellow,blue', it will search and download images for 'car red', 'car yellow' and 'car blue' individually

limit l

You can specify any integer value here. It will try and get all the images that it finds in the google image search page.

If this value is not specified, it defaults to 100.

Note: In case of occasional errors while downloading images, you could get less than 100 (if the limit is set to 100)

related_images ri

This argument downloads a ton of images related to the keyword you provided.

Google Images page returns list of related keywords to the keyword you have mentioned in the query. This tool downloads images from each of those related keywords based on the limit you have mentioned in your query

This argument does not take any value. Just add '--related_images' or '-ri' in your query.

format f

Possible values: jpg, gif, png, bmp, svg, webp, ico

color co

Denotes the color filter that you want to apply to the images.

Possible values: red, orange, yellow, green, teal, blue, purple, pink, white, gray, black, brown

color_type ct

Denotes the color type you want to apply to the images.

Possible values: full-color, black-and-white, transparent

usage_rights r

Denotes the usage rights/licence under which the image is classified.

Possible values:

• labeled-for-reuse-with-modifications,
• labeled-for-reuse,
• labeled-for-noncommercial-reuse-with-modification,
• labeled-for-nocommercial-reuse
size s

Possible values: large, medium, icon, >400*300, >640*480, >800*600, >1024*768, >2MP, >4MP, >6MP, >8MP, >10MP, >12MP, >15MP, >20MP, >40MP, >70MP

exact_size es

You can specify the exact size/resolution of the images

This value of this argument can be specified as <integer,integer> where the fist integer stands for width of the image and the second integer stands for the height of the image. For example, -es 1024,786

Note: You cannot specify both 'size' and 'exact_size' arguments in the same query. You can only give one of them.

aspect_ratio a

Possible values: tall, square, wide, panoramic

type t

Possible values: face, photo, clip-art, line-drawing, animated

time w

Denotes the time the image was uploaded/indexed.

Possible values: past-24-hours, past-7-days

time_range wr

Denotes the time range for which you want to search the images

The value of this parameter should be in the following format '{"time_min":"MM/DD/YYYY","time_max":"MM/DD/YYYY"}'

delay d

Time is to be specified in seconds. But you can have sub-second times by using decimal points.

url u

If you are searching an image on the browser google images page, simply grab the browser URL and paste it in this parameter It will download all the images seen on that page.

single_image x Allows you to download one image if the complete (absolute) URL of the image is provided
output_directory o

Allows you specify the main directory name in which the images are downloaded.

If not specified, it will default to 'downloads' directory. This directory is located in the path from where you run this code

The directory structure would look like: <output_directory><image_directory><images>

image_directory i

This lets you specify a directory inside of the main directory (output_directory) in which the images will be saved

If not specified, it will default to the name of the keyword.

The directory structure would look like: <output_directory><image_directory><images>

no_directory n

This option allows you download images directly in the main directory (output_directory) without an image_directory

The directory structure would look like: <output_directory><images>

proxy px

Allows you to specify proxy server setting for all your requests

You can specify the proxy settings in 'IP:Port' format

similar_images si

Reverse Image Search or 'Search by Image' as it is referred to on Google.

specific_site ss Allows you to download images with keywords only from a specific website/domain name you mention.
print_urls p

Print the URLs of the images on the console. These image URLs can be used for debugging purposes

This argument does not take any value. Just add '--print_urls' or '-p' in your query.

print_size ps

Prints the size of the images on the console

The size denoted the actual size of the image and not the size of the image on disk

This argument does not take any value. Just add '--print_size' or '-ps' in your query.

print_paths pp

Prints the list of all the absolute paths of the downloaded images

When calling the script from another python file, this list will be saved in a variable (as shown in the example below)

This argument also allows you to print the list on the console

Prints the metada of the image on the console.

This includes image size, origin, image attributes, description, image URL, etc.

This file can be found in the logs/ directory. The name of the file would be same as the keyword nam

socket_timeout st

Allows you to specify the time to wait for socket connection.

You could specify a higher timeout time for slow internet connection. The default value is 10 seconds.

thumbnail th

Thumbnails are saved in their own sub-directories inside of the main directory.

This argument does not take any value. Just add '--thumbnail' or '-th' in your query.

language la

Defines the language filter. The search results are automatically returned in that language

Possible Values: Arabic, Chinese (Simplified), Chinese (Traditional), Czech, Danish, Dutch, English, Estonian. Finnish, French, German, Greek, Hebrew, Hungarian, Icelandic, Italian, Japanese, Korean, Latvianm, Lithuanian, Norwegian, Portuguese, Polish, Romanian, Russian, Spanish, Swedish, Turkish

prefix pr

A word that you would want to prefix in front of actual image name.

This feature can be used to rename files for image identification purpose.

chromedriver cd

With this argument you can pass the path to the 'chromedriver'.

The path looks like this: "path/to/chromedriver". In windows it will be "C:\path\to\chromedriver.exe"

safe_search sa

Searches for images with the Safe Search filter On

And this filter will be Off by default if you do not specify the safe_search argument

This argument does not take any value. Just add '--safe_search' or '-sa' in your query.

no_numbering nn

When you specify this argument, the script does not add ordered numbering as prefix to the images it downloads

If this argument is not specified, the images are numbered in order in which they are downloaded

This argument does not take any value. Just add '--no_numbering' or '-nn' in your query.

offset of

This argument takes integer. Make sure the value of this argument is less than the value of limit

help h show the help message regarding the usage of the above arguments

Note: If single_image or url parameter is not present, then keywords is a mandatory parameter. No other parameters are mandatory.

## Config File Format

You can either pass the arguments directly from the command as in the examples below or you can pass it through a config file. Below is a sample of how a config file looks.

You can pass more than one record through a config file. The below sample consist of two set of records. The code will iterate through each of the record and download images based on arguments passed.

{
"Records": [
{
"keywords": "apple",
"limit": 5,
"color": "green",
"print_urls": true
},
{
"keywords": "universe",
"limit": 15,
"size": "large",
"print_urls": true
}
]
}

## Examples

• If you are calling this library from another python file, below is the sample code
from google_images_download import google_images_download   #importing the library

arguments = {"keywords":"Polar bears,baloons,Beaches","limit":20,"print_urls":True}   #creating list of arguments
print(paths)   #printing absolute paths of the downloaded images
• If you are passing arguments from a config file, simply pass the config_file argument with name of your JSON file
$googleimagesdownload -cf example.json • Simple example of using keywords and limit arguments $ googleimagesdownload --keywords "Polar bears, baloons, Beaches" --limit 20
• Using Suffix Keywords allows you to specify words after the main keywords. For example if the keyword = car and suffix keyword = 'red,blue' then it will first search for car red and then car blue
$googleimagesdownload --k "car" -sk 'red,blue,white' -l 10 • To use the short hand command $ googleimagesdownload -k "Polar bears, baloons, Beaches" -l 20
$googleimagesdownload --keywords "logo" --format svg • To use color filters for the images $ googleimagesdownload -k "playground" -l 20 -co red
• To use non-English keywords for image search
$googleimagesdownload -k "北极熊" -l 5 • To download images from the google images link $ googleimagesdownload -k "sample" -u <google images page URL>
$googleimagesdownload -k "boat" -o "boat_new" • To download one single image with the image URL $ googleimagesdownload --keywords "baloons" --single_image <URL of the images>
$googleimagesdownload --keywords "baloons" --size medium --type animated • To download images with specific usage rights $ googleimagesdownload --keywords "universe" --usage_rights labeled-for-reuse
$googleimagesdownload --keywords "flowers" --color_type black-and-white • To download images with specific aspect ratio $ googleimagesdownload --keywords "universe" --aspect_ratio panoramic
• To download images which are similar to the image in the image URL that you provided (Reverse Image search).
$googleimagesdownload -si <image url> -l 10 • To download images from specific website or domain name for a given keyword $ googleimagesdownload --keywords "universe" --specific_site example.com

===> The images would be downloaded in their own sub-directories inside the main directory (either the one you provided or in 'downloads') in the same folder you are in.

## Troubleshooting Errors/Issues

#~~~# SSL Errors

If you do see SSL errors on Mac for Python 3, please go to Finder —> Applications —> Python 3 —> Click on the ‘Install Certificates.command’ and run the file.

While using the above commands, if you get Error: -bash: googleimagesdownload: command not found then you have to set the correct path variable.

To get the details of the repo, run the following command:

$pip show -f google_images_download you will get the result like this: Location: /Library/Frameworks/Python.framework/Versions/2.7/lib/python2.7/site-packages Files: ../../../bin/googleimagesdownload together they make: /Library/Frameworks/Python.framework/Versions/2.7/bin which you need add it to the path: $ export PATH="/Library/Frameworks/Python.framework/Versions/2.7/bin"

When you run the command, it downloads the images in the current directory (the directory from where you are running the command). If you get permission denied error for creating the downloads directory, then move to a directory in which you have the write permission and then run the command again.

#~~~# Permission denied while installing the library

On MAC and Linux, when you get permission denied when installing the library using pip, try doing a user install.

\$ pip install google_images_download --user

You can also run pip install as a superuser with sudo pip install google_images_download but it is not generally a good idea because it can cause issues with your system-level packages.

#~~~# Installing the chromedriver (with Selenium)

If you would want to download more than 100 images per keyword, then you will need to install 'selenium' library along with 'chromedriver' extension.

If you have pip-installed the library or had run the setup.py file, Selenium would have automatically installed on your machine. You will also need Chrome browser on your machine. For chromedriver:

On Windows or MAC if for some reason the chromedriver gives you trouble, download it under the current directory and run the command.

On windows however, the path to chromedriver has to be given in the following format:

C:\\complete\\path\\to\\chromedriver.exe

On Linux if you are having issues installing google chrome browser, refer to this CentOS or Amazon Linux Guide or Ubuntu Guide

For All the operating systems you will have to use '--chromedriver' or '-cd' argument to specify the path of chromedriver that you have downloaded in your machine.

If on any rare occasion the chromedriver does not work for you, try downgrading it to a lower version.