Annotation Tool V 0.0.1

This is very simple annotation tool, with minimum 3rd party libraries, based on HTML/CSS/JS & Python (Flask).

For annotation manipulations used interact.js.

To generate image list place your images to static/img it's much better to keep unix type file names (snake_case without special symbols). Right now .PNG, JPG, JPEG formats are supported.

Make sure that data.json exists in the project root.

You can view detailed video presentation here.

One time setup

To run application you need one time setup per machine.

Make sure you've installed python 3.8 or higher.

Just open terminal for linux or cmd for windows and type python

Create virtual environment using script:

• For Linux Mac BSD systems setup.sh
• For Windows setup.cmd

Or you can do it manually:

https://packaging.python.org/en/latest/guides/installing-using-pip-and-virtual-environments/

Windows users:

python -m pip install --upgrade pip

python -m pip --version

Install virtualenv

python -m pip install --user virtualenv

And now go to project directory cd C:\annotation_tool

and create venv

python -m venv venv

Now activate your env:

.\env\Scripts\activate

Run application

All you need is run main.sh for Linux / Mac or click main.cmd for Windows .

Note for display any changes you need to rerun application.

Usage specifications

There is no image upload mechanism for this application. So to build annotation make sure that you have 2 components:

• In /static/img placed images
• data.json in project root

IMPORTANT please choose file names without spaces or special characters, the best way is shown in an example bellow:

file_name.extenstion

Supported file types

All images supported by browsers can be used, now it's tested with file types: .PNG, JPG, JPEG

Folder stricture

.
├── ...
├── static                           # Static folder files and folders located here are accessable for browsers
│   ├── img                          # !!!IMPORTTANT!!! Your content (IMAGES) must be here...
│   └── w3.css                       # Stylesheet CSS for goo visual
├── templates                        # Here are files which python renders in browser
│   └── index.html                   # Main file which is shown for user at start, most operations are called from here
├── tests                            # Test running via pytest engine
├── .gitignore                       # Requered to prevent load venv to git repository
├── app.py                           # Programmatically everything nice happens here
├── data.json                        # Here is stored annotations data, coordinates, other values including image names from img folder
├── LICENSE                          # Standard license from upwork copied here
├── main.cmd                         # Start application script for MS Windows users
├── main.sh                          # Start application script for Linux users
├── README.md                        # you're here
├── requirements.txt                 # Installed packeges via pip are stored here.
├── setup.sh                         # Setup script for Linux users
└── setup.cmd                        # Setup script for MS Windows users

How it works

All images are placed to statisc/img directory only image/png, image/jpg, image/jpeg are supported.

Python script will read this images, after saving image annotations program will save image name, and path to data.json file.