Image Toolkit v0.1.2

This is an image tool package for organizing photos. The main features are:

• Find visually duplicated photos: by comparing the photo hashes, the script can identify visually identical photos
• Find blurred photos: by calculating Laplacian value of the photos, the script can also identify blurred photos
• Remove fake transparent background: we use various image processing methods to remove fake checkerboard transparent background
• Remove duplicated images prefix: after processing the duplicated photos, there is a prefix added to the duplicated images; you can use our tool to undo the renaming.

Note: For blurry and duplicated photo checks, only JPEG images are tested and supported in this version. For fake transparent background removal, we only support PNG at this moment.

Basic Usage

Function 1: Identify Blurry and Duplicated JPEG Images

In the command line, you can just use the following command:

imgtoolkit

Blurred and duplicated image checks on the current folder will be automatically executed; blur and duplicate folders will be created to store blurry and duplicated images.

Sample Output

Start finding blurs
|████████████████████████████████████████| 12148/12148 [100%] in 18:38.0 (10.87/s)
11 blur photos processed, moved to blur/
Elapsed Time:  0:18:38.067728
Start finding duplicates
Phase 1 - Hashing
|████████████████████████████████████████| 12137/12137 [100%] in 2:25.1 (83.63/s)
Phase 2 - Find Duplicates
Phase 3 - Move Duplicates
93 duplicated images moved to duplicate/

Function 2 (Available since v0.1.2): Remove Duplicated Images' Prefix (after using Function 1)

To remove the prefix of the duplicated images in the duplicate folder, you can execute the following command by specifying target folder:

imgtoolkit remove_duplicate_prefix the_folder_which_needs_action/

Function 3 (Available since v0.1.1): Remove Fake PNG Background

To remove the fake PNG background, you can use the following command:

imgtoolkit remove_fakepng_bg source.png save.png

where source.png is the PNG that requires background removal, while save.png is the path you want to save the fixed image.

---

Using Our Toolkit in Python

In Python, you can find blurry and duplicated photos as follows:

from imgtoolkit import tools

if __name__ == '__main__':
  # Find blur photos
	tools.find_blur()

  # Find duplicated photos
	tools.find_duplicate()

Parameters used in Python

find_blur(folder='blur/', threshold=20)

• folder (string): the script will create a folder with a specified name and move blurred photos to the folder.
• threshold (integer): the lower the value, the stricter the check. The default value of 20 is generally okay for most cases.

find_duplicate(folder='duplicate/', prefix='DUPLICATED_')

• folder (string): the script will create a folder with the specified name and move duplicated photos to the folder.
• prefix (string): the script will append the prefix to the filename when moving to the duplicate folder. Set it to an empty string to avoid this behavior.

Installation

If you have installed pip, you can install this package via this command:

pip install imgtoolkit

Upgrade

To upgrade an existing installation, you can use the following command:

pip install imgtoolkit --upgrade

Questions or Suggestions

Please open an issue if you find problems using this script. Suggestions or feature requests are also welcomed.

Future Development

• More command line parameters

Donation

Donations are welcomed. We accept ETH or SOL.

• ETH: 0x5B3318109932c8EDc6297197895afDD68567672D
• SOL: GLBLjrYxhbncTFR6CbRTrpPjtaNJax13A82C8F3ceSAX