Extended Object Detection (EOD) is a computer vision library with idea to represent an object as set of attributes, and then more represent complex objects as sets of simple objects and relationships between them. For attributes detection, this library incorporates many famous other libraries, such as OpenCV, Dlib, Torch and so on. Other library idea is to provide configuration file, that could set up detection system, without diving in library's API. Solution is created for usage in robotics and also has some tools to process result of detection system.
Repository
@misc{extended_object_detection,
author = {Moscowsky, Anton},
booktitle = {GitHub repository},
publisher = {GitHub},
title = {{Extended Object Detection}},
url = {https://github.com/Extended-Object-Detection-ROS},
year = {2020}
}Article
@incollection{moscowsky_eod,
author = {Moscowsky, A. D.},
booktitle = {Smart Electromechanical Systems: Recognition, Identification and Modeling},
doi = {10.1007/978-3-030-97004-8_3},
keywords = {Computer vision,Object detection,ROS,Robotic},
pages = {27--43},
title = {{Extended Object Detection: Flexible Object Description System for Detection in Robotic Tasks}},
url = {https://link.springer.com/10.1007/978-3-030-97004-8_3},
volume = {419},
year = {2022}
}
- ROS Russia Open-Source Contest - 1st Place
- OpenCV Blog Olympics Contest - 1st Place (post)