Aruco tags detection & tracking performance on XR Elite

[wTrebol]

Hello,

 

For a proof of concept module we're making for a safety training, we want to make use of pass through in combination with spwaning certain objects on locations indicated by Aruco tags.

However when implementing the tag detection and trying the Unity Marker demo scene, we see that the performance is just awful and unusable.

• To detect a tag you have to look at it straight on for 10+ seconds 
• sometimes it just doesn't detect at all
• when looking sideways (not straight on) chances of the tag being detected are almost none
• we tried different sizes of tags, no difference
• same result with printed tags or displaying on a screen
• same results with office light dimmed or full on
• same problems when trying to track a detected marker.

Has anyone got this working with more decent detection? Is this module still in active development?

Thanks.

 

[ymiecielica]

Hi,

We have exactly the same problem. We'd like to use the ArUco as a common reference point for several headsets, so that they can all anchor in the same place, but the unreliability of the tracking makes this impossible.

Please can Vive do something about this.

 

Thanks

[Tony PH Lin]

Hi @wTrebol, @ymiecielica,

Sorry for the late response, and we did observe and receive some issues feedback from different sources and also internally on this beta version.

We will have an updating SDK and ROM version with improvement in next FOTA around late Sep.

In addition, we will also introduce Spatial Anchor feature in the same timeframe.

Will keep posted when the time is more closed.

Thanks.

[Maeve785]

1. The Aruco tags detection and tracking performance on XR Elite showcases impressive accuracy and reliability, enhancing user experience in extended reality (XR) applications. Leveraging advanced computer vision algorithms, XR Elite efficiently recognizes and tracks Aruco markers, facilitating seamless interaction and precise spatial mapping for immersive XR experiences. https://storysaver.page/

    

    

 

Edited April 1 by Maeve785

[nande]

Thanks Tony. Has this been implemented already?

[Tony PH Lin]

Hi @nande,

 

Yes, the Spatial Anchor is enabled from WAVE SDK 5.4.0 few months ago 2023.
https://hub.vive.com/storage/docs/en-us/ReleaseNote.html#release-5-4-0

Also we showcased in March GDC the multiple players with Mixed Reality using Spatial Anchor as coordination alignment. 

Check it out for the video, and we're planning as open source project on Github shortly.

 

[nande]

Thanks a lot Tony, and thanks for the video.

If possible please reply here once the opensource sample project is ready 🙏

thank!!

[Nage1952]

On 8/6/2023 at 2:22 PM, wTrebol said:

Hello,

 

For a proof of concept module we're making for a safety training, we want to make use of pass through in combination with spwaning certain objects on locations indicated by Aruco tags.

However when implementing the tag detection and trying the Unity Marker demo scene, we see that the performance is just awful and unusable.

• To detect a tag you have to look at it straight on for 10+ seconds 
• sometimes it just doesn't detect at all
• when looking sideways (not straight on) chances of the tag being detected are almost none
• we tried different sizes of tags, no difference
• same result with printed tags or displaying on a screen
• same results with office light dimmed or full on
• same problems when trying to track a detected marker.

I will try to solve this problem on my own, but as a last resort I have the option of turning to [url= https://www.cogniteq.com/blog/top-iot-companies] iot companies[/url] to develop special software for myself or fix a problem in existing code with the help of the best professionals in the business.

Has anyone got this working with more decent detection? Is this module still in active development?

Thanks.

 

It looks like you're having some serious issues detecting tags in the Unity Marker demo scene. While Aruco markers can be useful for augmented reality applications, they can actually cause performance issues under certain conditions.

Here are some suggestions and thoughts:

Optimize marker detection settings. Make sure you configure your marker detection settings in Unity to optimize performance. This may include adjusting thresholds, marker sizes, or other parameters to improve detection reliability and speed.

Check lighting conditions. Inconsistent or poor lighting conditions can significantly affect marker detection. Experiment with different lighting settings to see if there is an optimal configuration that improves detection rates.

Explore alternative marker systems. While Aruco markers are popular and widely used, they may not be the best solution for every scenario. Consider exploring alternative marker systems or tracking methods that may provide better performance for your specific application.

Update your software and libraries. Make sure you are using the latest versions of Unity, Aruco and any associated libraries or plugins. Sometimes software updates include performance improvements and bug fixes that can solve problems like the ones you're experiencing.

Community support and forums. Reach out to the developer community for support and valuable information. Forums, online communities, and developer groups dedicated to Unity and AR development can offer valuable advice, troubleshooting tips, and potential solutions based on the experiences of others.

Consider customized solutions. Depending on the specific requirements of your security training module, you may need to explore custom solutions or additional developments to achieve the desired performance and functionality.

Regarding the development status of a module and whether it is still actively supported, it may be helpful to contact the developers directly or review their documentation and release notes for any updates or announcements.

I hope these suggestions help improve tag detection performance in Unity. Feel free to experiment with different approaches and seek help from the developer community to overcome these problems. Good luck with your safety training!