Image tracker on iOS issues

🌵 Needle Engine
1

Greetings,

I am currently experimenting with the image tracking sample. On Android, it works very well and detects the marker accurately. However, I am encountering issues with iOS. When the iPhone points at the marker, the model appears at a very small scale. This issue occurs even in the sample from the official website in QuickLook. We tested with iPhone 15, 16 Pro Max, 12, and 14 Pro, and all of them suffer from the same problem. Additionally, the model sometimes forces itself to open as a plane tracker. I am attaching an image to show how the model looks on iOS.

image tracking issue
image tracking issue590×1280 93.2 KB

2

Hello Camilo,

The issue you’re experiencing with image tracking on iOS, where the model appears at a very small scale, can often be related to how AR image tracking is set to interpret physical sizes. Here are some key aspects to troubleshoot and consider:

  1. Check the Physical Width: Ensure that the widthInMeters property of the tracked image is set correctly. This value should accurately represent the physical size of the image being tracked.

  2. Scaling Adjustments: In WebXRImageTracking, there may be code that applies scaling factors when positioning the tracked model. Make sure the scaling is set correctly, especially in the portion that applies the transformation from camera coordinates to world coordinates (see the documentation).

  3. Model Size Configuration: Review how your model’s dimensions are defined. If it’s a GLTF or USDZ file, improper scaling can lead to unexpected sizes. Ensure that the model has appropriate dimensions relative to the tracked image’s physical size.

  4. Debugging with Console Log: Introduce some debugging logs within the image tracking update functions to see values being calculated for positions and scales. This can help identify where scaling may be going wrong.

  5. Multiple Images: Confirm that you are only tracking one image at a time, as QuickLook has limitations regarding multiple tracked images, which may lead to unexpected behavior (see the documentation).

  6. Plane vs. Image Tracking Confusion: Sometimes if the model incorrectly detects itself as a plane tracker, it might affect how size transformations are applied. This can occur based on how your anchoringType is set (see the documentation).

  7. Testing Different Models: It might be wise to test with different marker images or models to see if the issue persists across the board, giving insights into whether the problem is model-specific.

I recommend verifying these aspects in your implementation and testing accordingly. If you continue to encounter issues, providing specific snippets of code related to your image tracking setup would assist in diagnosing the problem further.

Feel free to ask if you need further assistance!

If you have additional details or specific code examples to share, that would be helpful for a deeper analysis.

1 replies left - Get a PRO license for more AI replies.
thread_Q94VKIh1rDRC8zbNaTFT35Yy
Reply to this message to continue the conversation