YOLO Ultralytics Info

[ElDablo123]

For our object detection using YOLO once we have arrived at the first GNSS coordinates we have to complete challenges such as find our ArUco posts. The GNSS coordinates for each of these posts are not very accurate ranging from 5 - 20 meters from the actual location.
The posts have the ArUco tags on 3 sides of the post, so for training a model to identify these especially from far away we should build our own post and create our positive and negative images to train.
Exact ArUco description info based on rulebook:
"The posts and gate will have 3-sided visual markers with 20 x 20 cm faces, 0.5 – 1.5m off the ground. Each marker will display a black and white
ARUCO tag using the 4x4_50 tag library exactly as shown below. Each tag will have a white border 1 cell in width, so cells are 2.5 cm across. The same tag will be displayed on each of the 3 sides so that it is visible from any direction."

Next we also want to identify objects that are on the ground with GNSS coordinates of around <10 meter accuracy however we still need to be within 2 meters to complete the challenge.
These objects are an orange rubber mallet and a plastic water bottle with no exact photo but just a description of "The second object will be a standard 1 liter wide-mouthed plastic water bottle (approximately 21.5 cm tall and 9 cm diameter)
and of unspecified color/markings color (not the same bottle used at URC2024)."

To train these models we firstly need a replica of the object we want to identify and several pictures annotated. To actually go and train the model may take a lot of time and power so I'd recommend learning Google Colab to help with the training.