Keypoint Matching Issues and Multiple Blocks

[levon]

Hello Team, I am currently processing a project in which we had 3 different flight missions. Flown at a 75 overlap at 200’ AGL. We are currently using a Zenmuse P1 35mm and M300 w RTK. I have tried processing the 2600+ images all at once, and as you would expect this created multiple blocks within the data. We have 5 GCPs across the project and I am looking for some direction in trying to solve these processing issues.

Processing Options
General - 1/2 Image Scale
Matching - Aerial without G.V.M.
Calibration - Auto, Standard, Internal Camera Parameters set to All Prior, Rematch Auto

My current process is separating the individual flight missions and processing them separately. Adding MTPs between project and in the process of attempting to merge them. That’s when I noticed however even one of 2 of the flights are creating multiple blocks as well. My understanding is that this was just poor data collection, probably could have used a higher overlap as well as consistent flight times. Or could this be an issue with the P1 and camera? Any help, information, or knowledge would be appreciated.

Reports
12062-1_report.pdf (1.5 MB)
12062-2_report.pdf (1.5 MB)
12062-3_report.pdf (3.4 MB)

[Mike_K]

Hi levon,

Trying to map areas with heavy vegetation is challenging. It has a complex geometry that can move in the wind. As a result, it makes for a challenging environment. In these areas, you will want to have a very high overlap. Probably between 80-90% front and side.

You can also drop the image scale in step 1 down to 1/4. This tends to reduce the amount of noise caused by the veg and will help with the calibration. Give this a shot.

I would also change your calibration method to Alternative. This is designed for agricultural applications but works well in vegetation. You can also try the Accurate Geolocation and Orientation as well.

[josh.abbott]

Hello, i also have 3 blocks showing up on a beach survey recently completed. I cannot locate these different blocks to try fix. Can someone assist on how to identify where the 3 blocks are?
It is creating multiple height layers in a particular area of the survey.

[josh.abbott]

Ashburton_May25_Zone7run3_report.pdf (1.2 MB)

[Mike_K]

Hi Josh,

The quality report looks very good. The presence of three blocks may seem concerning, but it’s a bit misleading. The screenshot below highlights their locations:

• The yellow block represents a single, continuous area that covers the majority of your project.
• The red and green blocks are small, isolated sections near the water.

[Mike_K]