A Verification of AprilTags for a Pharmacy Application

Moore, Cassandra (2020) A Verification of AprilTags for a Pharmacy Application. [USQ Project]

Text (Project)

Download (9MB) | Preview


AprilTags are an artificial tag that provides pose estimation. Commonly, it is used to determine global coordinates of a space so that robots can determine where they are within the area. In this project, however, the AprilTags were attached to a specific object needing to be moved. The scope was to verify the robustness and functionality of the tags in order to determine the plausability of them being used for a Pharmacy application.

The Pharmacy is a highly crucial part of the hospital system and the time of the Pharmacists is of the upmost importance. One of the tasks of a Pharmacist is to unload the boxes of medication from delivered orders and organise it into the dispensary. Robotic arms already exist to help with the retrieval and stocking in the dispensary. However, the process of sorting the delievered orders has not been automated and this is a timeconsuming task for a Pharmacist or Pharmacist assistant. Hence, the thought towards verifying the functionality of AprilTags was to determine whether they could be used on medication packaging to potentially automate the process of unboxing.

The method undertaken to verify the tags took on a couple of stages. In each stage, an OpenMV H7 Camera was used for the detection process. This camera was attached to Haddington Dynamics’ Dexter HDI robotic arm. An AprilTag was attached to the test object and tests were run to verify the tags ability to handle various distances between the camera and tag, the size of the tag, the angle of the camera, occlusion to the tag, and lighting.

The results of these tests were interesting. It appeared that there was only a limited range of distances, tag sizes, camera angles, and occlusion that could still result in a detection accuracy above 80%. The smallest tag that gave functional results was the 25mm tag. The optimal conditions for high accuracy was the camera being between 10 and 20 cm above the tag with a camera angle of 90 degrees, with controlled lighting and no more than 17% occlusion.

From the results and analysis, the tags are a possibility for automating the unboxing of medication packages. Further testing needs to be conducted with the pose estimation to determine its accuracy. Other detection methods may need to be used alongside the AprilTags since this project’s testing has demonstrated the limited range from where the tag can be detected.

Statistics for USQ ePrint 43044
Statistics for this ePrint Item
Item Type: USQ Project
Item Status: Live Archive
Faculty/School / Institute/Centre: Historic - Faculty of Health, Engineering and Sciences - School of Mechanical and Electrical Engineering (1 Jul 2013 - 31 Dec 2021)
Supervisors: Low, Tobias
Qualification: Bachelor of Engineering (Honours) (Mechatronics) / Bachelor of Science (Mathematics)
Date Deposited: 16 Aug 2021 05:25
Last Modified: 26 Jun 2023 04:44
URI: https://sear.unisq.edu.au/id/eprint/43044

Actions (login required)

View Item Archive Repository Staff Only