imageSimilarity

imageSimilarity is a tool that divides a group of images according to their similarity grade between them which goes between 0 and 1. 0 means not similar and 1 means that those images are equal. It uses the SSIM method. Just a interesting fact, the relation of similarity is considered to be transitive, symmetric, and reflexive. It's a key concept utilized when images dividing comes.

> python .\imageSimilarity.py
     _____
    |_   _|
      | | _ __ ___   __ _  __ _  ___
      | || '_ ` _ \ / _` |/ _` |/ _ \
     _| || | | | | | (_| | (_| |  __/
     \___/_| |_| |_|\__,_|\__, |\___|
                           __/ |
                          |___/
     _____ _           _ _            _ _
    /  ___(_)         (_) |          (_) |
    \ `--. _ _ __ ___  _| | __ _ _ __ _| |_ _   _
     `--. \ | '_ ` _ \| | |/ _` | '__| | __| | | |
    /\__/ / | | | | | | | | (_| | |  | | |_| |_| |
    \____/|_|_| |_| |_|_|_|\__,_|_|  |_|\__|\__, |
                                             __/ |
                                            |___/
    
    Loading images from 'images'...
    Dividing images in batches according to similarity (0.95)...
    Saving images in 'output'...
    Done.

1. Contents
2. Prerequisites
3. Running
4. Run with docker
5. Development and Debug
6. APIs
7. Get Help

Getting Started

To get the repository, run:

 git clone https://github.com/codexgigassys/imageSimilarity.git

Prerequisites

To install all Python (3.9 <= ) required modules, please run:

 pip install -r requirements.txt

Running

First, you need a directory with images. You can save them inside images directory or you can create your own folder outside the script main folder. In case you do the later option, when you run the script, you pass the argument --imagesDirectory pathToImages (it works with relative and absolute paths).

Once the script is done, you have the result inside output directory or you can create your own folder passing the argument --outputDirectory pathToSaveOutput. If the folder doesn't exist, the script will create it.

Remember that similarity grade goes between 0 and 1. By default, this value is 0.95. In case you want to use another value, pass the argument --similarityGrade valueBetween0And1.

A complete script running is:

python .\imageSimilarity.py --similarityGrade aValidValue --imagesDirectory pathToImages --
outputDirectory pathToSaveOutput 

If you wish to log imageSimilarity activities, add --logging or just -l. All logging related information may be in imageSimilarity.log.

Run with Docker

All you have to do is

docker-compose up

API of below will be working

Development and Debug

In case you want to debug or make a contribution to Image Similarity, you can run docker-compose -f build-n-debug-yml up -d. This command will mount an image-similarity volume in your computer, allowing you to develop or debug with simplicity.

APIs

The script can create a server to hold an API that receives images and then returns a list of lists with similar images hashes. A use case could be (it's exampleCode.py):

import requests
import json

filesDictionary = {
    '01.jpeg': open('images/01.jpeg', 'rb'),
    '02.jpeg': open('images/02.jpeg', 'rb'),
    '05.jpeg': open('images/05.jpeg', 'rb'),
}

response = requests.post("http://localhost:9082/api/imageSimilarityByHash",files=filesDictionary,params={'similarity_grade':0.75})

hashes = json.loads(response.text)['lists']

If you print hashes you can obtain:

[["52d57c19cc025fe44827e0970fa35d8ba5a5441f", "39e25885c9c11ab8299668569c1162ff610868e9"], ["4dedb7caab44c5b65ba0692cc54a5f2735dee5f4"]]

If you don't pass 'similarity_grade', it will be 0.95.

To run the server, just type:

python .\imageSimilarity.py --runServer

Get Help

To get help:

python .\imageSimilarity.py --help