History

Otary was at the beginning only a part of a larger project in a private repository started in 2024.

This private project was about processing architectural cadastral documents. It required segments detection, contours detection, corners detection, document layout analysis, optical character recognition (OCR), custom Deep Learning Computer Vision models and much more.

I needed a general image, geometry and vision processing library that would be able to help me at a low-level for all my Computer Vision tasks.

At the start, I was just tired of juggling with all the python image libraries to perform both basic and specific image processing tasks. I just wanted something that was simple to use and easy to understand.

I needed a tool that combines image, geometry and compute vision processing, this is where the journey started.

I decided to make Otary public and open source in July 2025.

This is how Otary was born.