resolving uncertainty ellipses and multiple blocks

[bla]

I have a field that I’ve flown multiple times this year using a Mavic 3 Multispectral with a D-RTK 2. 80% front overlap, 70% side overlap (pilot 2 defaults) at 85’. All previous flights with either camera have produced no/negligible uncertainty ellipses and a single block. The most recent flight looks as expected for the multispectral images, but produced several uncertainty ellipses and 3 blocks for the rgb images. Both rgb and multispectral images are captured simultaneously on the same flight. I’ve attached the quality reports for both the most recent rgb and multispectral flights.

I’m trying to determine if there is a way to improve the rgb results. I’d also like to make sure I’m using optimum flight settings (overlaps, etc…) for this type of field moving forward. Do you have any suggestions to improve the results?

midville_peanut_06-18-24_rgb_report.pdf (1.3 MB)
midville_peanut_06-18-24_ms_report.pdf (1.3 MB)

[Mike_K]

Hello,
I looked at your RGB quality report and have some ideas and suggestions. First, 85’ is a relatively low altitude at which to fly, resulting in a very small ground sampling distance (GSD). This small GSD can be problematic in vegetation. The problem is that vegetation has a complex geometry that can also move in the wind. The result is that an insufficient number of keypoints will be generated. Hence, you had large uncertainty ellipses and image blocks.

With your current dataset you can drop your image scale under step 1 to 1/2 or lower. This will tend to negate the effects of vegetation.

Typically, the best solution is to fly a little higher. A little larger GSD over vegetation should help. Furthermore, each image’s field of view will capture more terrain, ideally generating more common keypoints between images.