Reading RPLS is free for the whole profession. Members post, reply, and get the members-only rooms.
Last week we had over a foot of lake effect snow, followed by rain and warm temps. Today was the first day of calm winds and no rain, no snow. But still no sunshine. Took a chance and flew this with overcast skies, 26å¡F, with 1" of snow on ground. It turned out great. Look how easy it is to digitize the pool, sidewalks, and driveways. This is from a 10 minute flight time, and covers area 20 acres. Nice and warm now in the office to do the drafting.
This image has GSD 0.86 in or 2.19 cm.
Here is a look at the Pix4D Quality Report.
If you wouldn't mind...a few questions?
Where are your ground control points located on the image?
How many on the ground check shots were taken to verify the overall accuracy of the photo/model and what method was used to check it (total station or GPS)?
On that site I have 4 GCP's, 3 check points measured by with RTK and EDM, they are about 300ft apart in a zig-zag pattern from end to end.
I also have 80 EDM locations on building corners and fences.
The relative error is shown overly too accurate, beyond my ability of true measurements. As shown below the software derives an "Error to GCP Initial Position: -0.001ft, -0.000ft, 0.003ft". This is derived from me hand snapping to the GCP on 5 images. There is no way I can see the nail head the photos below.
Impressive! And it looks fun & satisfying. Is it really 167 images? Does the software automatically clip the edges where it can't model to your spec? Why do 3 of the GCP points look identical?
Baja,
Yes, 167 images were recorded in the 8 minute flight traveling about 11MPH at 180ft AGL.
Shown is the same GCP on four different images.
Lot's of fun right now. I get to draw these features in the office with mouse, rather than tramping through the snow, and wind with a prism pole. Sure glad I flew this yesterday. We got 4" of fresh snow overnight, and another 3"-5" on the way. We are on the east end of Lake Ontario, so the Lake Effect machine is running. Just 5 miles south, it's raining with no snow on the ground.
Yes, the software does clip the orthomosaic based on pixel matching and accuracy. I actually four parallel paths create a rectangle.
Local Government in one of the rare moments of good judgment did an aerial mapping/Lidar project that covers a fair portion of my work area in 2011. The aerial ortho photo tiles are georeferenced to GPS/CORS control and match up rather well with my design surveys. Granted resolution is only 0.5'-1'/pixel but close enough to allow me to do the same thing in the dead of winter when it's dark, there's 3' of snow on the ground, it's -20^ out and wind is howling. I still have to go out and tie down the critical improvements and tie in control, but sure beats chipping ice for an edge of pavement shot with cars and trucks whizzing by and it's plenty good enough for a lot of the rough design stuff I do.
Just because I'm paranoid, doesn't mean they aren't out to get me.
Now put the point cloud under this in C3D and you can draft with real elevations as well!
tomchurch, post: 403115, member: 10174 wrote: Now put the point cloud under this in C3D and you can draft with real elevations as well!
Tomchurch,
You got the right idea, except the point clouds works even better with Microstation and TopoDot software.
I've been doing the same thing. What benefits are you getting from the paid version of Pix4D? What error are you getting in your verticals?
leegreen, post: 403151, member: 2332 wrote: Tomchurch,
You got the right idea, except the point clouds works even better with Microstation and TopoDot software.
Yeah, I'm about to buy a solution for point cloud drafting and I'm leaning towards TopoDOT. We need to learn Microstation anyways. Also looking to try PointSense but I can't get in touch with them for a demo.
Sent from my SM-G935V using Tapatalk



