Hi,
I have a project where I have a fixed anchor and a mobile object; I need the precise straight line distance between these two points. They would be 10cm to 1m apart indoors with no obstructions. I've been looking at ultra wide band sensors, but they all seem to have accuracies in the 10cm range. If that's the best that I can do, how much do the distances fluctuate? With minimal movement could I expect consistent measurements? Maybe I'm thinking about this incorrectly and I could use another technology other than UWB sensors. My backup is to use a camera tracking system, but that is not ideal. Any feedback or discussion would be great.
For the UWB radio modules, within the range of +/- 10 cm. But that requires careful calibration of the modules. Average a bunch of (calibrated) measurements to get more precise values.
I suggest to use a different method of measuring such short distances. For specific suggestions, take the time to describe the setup in sufficient detail.
Thank you for the feedback. For specificity, I am trying to measure the distance from a user's headset (headphones or something wearable) to a display monitor. Shouldn't be too much movement, but there will be some.
You forgot to mention the desired accuracy and precision (not the same thing), and your description of the problem is still too vague to be very helpful.
UWB modules won't be much use, if any but there are several other non-contact measurement techniques to try (IR distance sensors, Lidar and TOF modules). Pololu has a good selection.
1 Like
Thank you for your patience despite my subpar specifications. I'll try be as detailed as possible.
In this case, precision is much more important for my application: I am looking to find the position of a user's head from a display monitor. This is an eye tracking project, and I need position data to train my models. My original idea was to use distance as a proxy for actual 3D position data, but of course 3D position data would be preferable. The 3D position route would involve recording the positions of a couple facial points, and the distance route would involve measuring the distance from one facial point to the sensor mounted on the monitor.
I would hope for a precision of less than a centimeter. Accuracy could be anything less than 20cm. Precision is far more important since my model needs to be able to recognize correlations rather than explicitly computing vectors with precise accuracy measurements.
I've noticed that my iPhone's new LiDAR sensor would provide this kind of 3D position data that I am looking for, so I've been looking into that avenue today. Stereo cameras might also work as well. The smaller the set up, the better, since I am only using appearance based eye tracking which is typically not used in conjunction with complex external hardware.
Eye tracking in the context of a user sitting in front of a screen sounds like a very difficult problem. There is much more to it than knowing the distance of the head to the screen, which was the original question, so I doubt I can help.
Yes, I was not looking for help regarding the entire problem, just sensor recommendations for accurate distances. Thanks for your help though.