Spatial Error and GCPs

[Reuben]

Using the GCPs in the screenshot below, I used GCP02, 06, 09, & 11 as controls and the rest as control + constraint.

1) What does the Error (m) mean? Is this horizontal, vertical, or a 3D error? What is the confidence of the error... 1-sigma?

2) For GCP01 my GNSS errors were 3.3 m horizontal and 5.2 m vertical (RMS 63-68% confidence). How do I compare that with the 4.3518 m error after scaling?

3) Also, the GCPs that are only controls (e.g. GCP02) has N/A. How do I assess the error of these control GCPs in meters?

[Andrea Alessi]

Hi Reuben,

sorry we didn't reply earlier to your thread - it fell below other and i missed it, our apologies!

1. the error is in meters (as you are using WGS84). It's a reprojection error, so to answer your question, it's 3D. What do you mean exactly with the error confidence?

2. i'm not sure i understand your question, but in general, your device error should be added to SfM error. Those errors looks very high, is this a flight at which altitude?

3. when N/A is shown, then you didn't insert in zephyr the coordinate (unless it's a bug

[Reuben]

Hi Andrea,

no worries! I understand this a bit more. The only thing is the confidence of the error of the GCP in Zephyr after scaling. What is the standard deviation of the error... would that be 1-sigma (68% confidence), or 2-sigma (95% confidence) for example? Or is it a direct measurement so it is 100% confidence?

cheers.

[Andrea Alessi]

Variance confidence could be calculated on one pixel only or depending on the sampling. We do not provide this information yet, if you can give us the formula you need to compare the confidence to, we can easily add it