Welcome to the ATHENAKY+ repository!

This repository was used to develop a novel Algorithm for Tracking using a Hybrid ENhanced Analytics Kalman + YOLO (ATHENAKY+) approach.

The database was acquired using a GEPRC CineLog35 drone equipped with a GoPro Hero 11 Black to capture the images. The image database used for training and cross-validation consisted of 5605 images, of which 1439 (25.67%) were negative samples. The images were acquired at the following resolutions: 1920 by 1080 and 640 by 480 pixels.

The database contains a recorded quadcopter and three different aircraft (a Cessna model, a Bixler V2, and a glider).

The DarkMark framework was used for image annotation. As a direct result, the database contains 2525 annotations with quadcopters and 1943 with planes—in this mode, there are 1.1 marks per image.

Another set of 1524 images was used solely for testing. Since the tracking algorithms developed require a sequence of images (they need prior detection to predict the information in the next frame), this database includes images from six videos. The images were numbered in ascending order and correspond to the temporal unfolding of the associated videos.

Note:

1. The ATHENA database is exclusively accessible for non-commercial research endeavours.

2. You agree not to reproduce, duplicate, copy, sell, trade, resell, or exploit for any commercial purposes any part of the images or any derived data.

3. You agree not to copy, publish, or distribute any part of the ATHENA database. However, making copies of the database is permitted solely for internal use within a single site of the same organization.

4. All submitted papers or any publicly accessible texts utilizing the ATHENA database are required to cite the following publication:

Link to the full paper: https://www.mdpi.com/2226-4310/12/8/649