Image Tracking Image requirements

Include the following details while filing a bug report (edit as applicable):

  • Issue category: Image Tracking
  • Device type & OS version: Any
  • Host machine & OS version: Any
  • Issue Environment : On Device
  • Xcode version:
  • ARDK version: 2.4.1
  • Unity version: 2021.3

Bug reproduction steps:

I am now getting extremely frustrated with the Lightship image tracking solution as it seems very nondeterministic as to what makes a “good” tracking image.

  • Images that pass ARCores image test do not work 100%
  • Images that were usable under ARFoundations Image tracking do not work with lightship

The documentation is purely throw away and does not give proper advice/guidance as to what makes a good trackable image:

  • No size requirements
  • As discussed previously, the mandated jpg requirement (only inferred from the code use)
  • No quality requirements

This image from the ARcode SDK works

This generated image does not (but works with ARcore and ARFoundation

All that is stated in the docs, which is totally insufficient is:


Images To Use

The quality of image tracking is highly dependent on the image being tracked. What makes an easily tracked image isn’t immediately intuitive. You’re looking for an image that:

  • Has many image features. Areas of high visual contrast. Something solid color or with a gradient won’t have the features you need.
  • Doesn’t have repeating patterns or features. A chess board has sharp visual features at the intersections of squares, but the visual features repeat constantly making them less useful for placing the image.

ARCore and ARKit both provide tools for determining the quality of a given image. You should use those tools if you need a way to verify an assumption about whether an image will be good or bad for tracking.

Can we please have accurate requirements or a way to test images that is actually reliable, I have now lost over half a day trying to find an issue with the images not being loaded, only to find it was because they were not accepted

See also:

Which talks about the issue with InvalidReferenceImage error which only tells me “something” happened.

Thank you for letting us know about your experience. Just wanted to let you know progress is ongoing in making improvements to the platform and particularly to image detection capability. I’ve notified our engineers of the demand for more feedback on whether a submitted image is good or bad. Please keep an eye on release notes as new updates come out and I will post you with progress on this feature