Image Categorizer

Description

These CLI programs help speed up and improve the manual image categorization process by providing a simple, but effective workflow.

Table of contents

• Image Categorizer
  • Description
  • Table of contents
  • How to use
    • Manual Binary Categorizer
      • Description:
      • Commands:
      • CLI Parameters:
      • Example of usage:
    • Manual Multiclass Categorizer
      • Description:
      • Commands:
      • CLI Parameters:
      • Example of usage:
    • Instalation
  • Technologies
  • Autor

---

How to use

Manual Binary Categorizer

Description:

CLI application to classify all images from a specific directory into two categories: A or B.

By calling the CLI application you will define the directory associated with A and B. When a image is classified, it is automatically moved from the source directory to the target directory assigned.

Commands:

• a: Move the shown image to directory A;
• b: Move the shown image to directory B;
• q: Exit the program;
• h: Print all commands and instructions in the terminal;

CLI Parameters:

• images_dir:
  • Source directory containing all images that should be classified.
  • Kind: Positional argument.
• -a or --dir-a:
  • Directory associated with the category A.
• b or --dir-b:
  • Directory associated with the category B.
• --create-dir:
  • If used both ,dir-a and dir-b, are going to be created if already do not exist.

Example of usage:

$ python manual_binary_categorizer ./dataset/images -a ./target/classA -b ./target/classB --create-dir

---

Manual Multiclass Categorizer

Description:

CLI application to classify all images from a specific directory into various categories (Maximum of 34 classes).

By calling the CLI application you will define a list of output directories, corresponding to each class you want to classify. This list o directories will be associated with all digits and lower-case letters, in that order. When a image is classified, it is automatically moved from the source directory to the target directory assigned.

Commands:

• digits and letters: Move the shown image to directory associated;
  • The directories listed are going to be assigned first by numeric order from [0,9] and then in alphabetical order [a, z].
• q: Exit the program;
• h: Print all commands and instructions in the terminal;

CLI Parameters:

• images_dir:
  • Source directory containing all images that should be classified.
  • Kind: Positional argument.
• -d or --dir-list:
  • Output directories separated by space.
• --create-dir:
  • If used both ,dir-a and dir-b, are going to be created if already do not exist.

Example of usage:

$ python manual_multiclass_categorizer ./dataset/images -d ./target/classA ./target/classB ./target/classC --create-dir

Instalation

All requirements are in the requirements.txt file. From pip, just run in the terminal the following command and you are ready to go.

$ pip install -r requirements.txt

Technologies

• Python>=3.8
• OpenCV>=4.5

Autor

Diogo Nascimento