Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.

Messages - kgrover

Pages: [1] 2 3
General / Local to projected coordinate system of point cloud
« on: July 20, 2023, 12:58:19 AM »
I have some existing point clouds as las / laz that I have imported into Agisoft, and imported as a local coordinate system (as we weren't sure of the source initially). I also have a CSV of control that was imported (as local), and they match to my point cloud.

But now I want to assign the projection of the point cloud (but not actually tranform the dataset) to a UTM Zone that we use, but when changing the Coordinate System in the reference settings to UTM (or really any chosen projection), it seems to be okay for the Control and just changes XYZ to NEElev, but is completely transforms the point cloud to a different location.

Is there a step missing to do this as other software does??

General / Re: Look through image view shifted from tie points / control
« on: January 06, 2020, 10:46:06 PM »
Hi Alexey, I was able to add in the images to my initial post.


General / Look through image view shifted from tie points / control
« on: December 23, 2019, 06:59:01 PM »
I have a real odd dataset right now that I can't figure out. It is of a runway collected by a sub using an Inspire and X5S camera. The images align fine, but they seem to be shifted / twisted from the tie points that are created (horizontally) and my survey control and I don't know why. They are all nadir imagery, so it isn't an oblique imagery issue at all.

The attached image shows the collected imaged on the right, and the look through of the same image on the left (with an orthographic view to show where the image is showing up. It appears the image is over the edge of the runway, but in fact it is much closer to the center of the runway. I have never seen a shift like this in all the projects I have collected over the years.

Also when I zoom out to just the point cloud with image locations, it appears all the image locations are shifted from the tie points, but in fact, they should be centered directly over the tie points.

I have tried to replicate with other imagery but can't do so, so I have no idea what is happening but open to suggestions.

**Note: i can't upload attachments for some reason at this time**

Feature Requests / Re: Background Map
« on: October 18, 2019, 04:02:14 PM »
So this is good for the DEM / Ortho, but the preference would be for the background map to be displayed in the model view as the main need. This way when images are imported and even reprojected, we can validate that it is being done properly. Same with when control is imported, all this is done in model view so having the background map there would be preferred.

It is almost too late in the game to have the background map at the DEM / Ortho view stages, as all the project info is typically set prior to this.

Feature Requests / Re: GCP vs Check Point Icon
« on: October 18, 2019, 03:58:57 PM »
Hi Alexey,

Primarily in model view when I am marking the control points, check shots, etc. So when I go thru the process of choosing which ones to hold as control and others to use as check points, they appear different in the model view (just to ensure I have proper placements / constraints visibly).

Feature Requests / GCP vs Check Point Icon
« on: October 18, 2019, 01:05:55 AM »
Any chance a different color of icon could be used to differentiate between a GCP and a Checkpoint? For big projects, when trying to set which ones to constrain as a GCP  would be nice to see a color (or different icon) between the two.

Trying to manually look thru the list of which are which is a bit awkward.

Feature Requests / Re: Background Map
« on: October 18, 2019, 12:58:19 AM »
Just wanted to bump this request up again. Would be extremely helpful.

Feature Requests / Re: Limit box for viewing / classification
« on: December 20, 2018, 10:56:55 PM »
I have been using that tool, which is good, but the problem is that you have to keep resetting the filter when you need to select a new area of points. It works (and a good first step), but kind of a pain with larger edits.

Ideally a limit box (similar to many other point cloud tools), pretty much like region box in Agisoft, but only that only controls the current view of points. So you can dynamically edit the limit box as needed (move, resize, etc.), instead of selecting a new area, filter by selection, the reset, and keep doing over again.

Feature Requests / Limit box for viewing / classification
« on: October 12, 2018, 08:26:58 PM »
It would be great to have the ability to set a limit box in the 3D view to better work with the dense cloud. It can be hard to properly select / view points that you need to clean, classify, etc. in the view as all the points are visible.

But with a limit box that you can resize move, etc. on the fly to clip the dense cloud would be ideal. Then it would allow for easier point selection of noisy points, or when manually classifying trees, buildings, etc.


Feature Requests / Re: Background Map
« on: August 15, 2018, 10:31:33 PM »
+1 for this as well. Bing, Google, Mapbox, it really doesn't matter. Just general context imagery would be great.

Feature Requests / Re: ECW export
« on: August 15, 2018, 10:30:22 PM »
+1 from here as well. We tend to convert exported tiffs to ECW in Globalmapper for better handling of data in CAD, GIS, etc. Directly in Agisoft would be awesome.


General / Re: Slow alignment time - compare between computers
« on: October 04, 2016, 09:25:27 PM »
Hi Alexey, any update or feedback on this??

I am still having the same issues with this dataset and others.


General / Re: Slow alignment time - compare between computers
« on: September 21, 2016, 07:22:41 PM »
Hi Alexey,

I have attached the log files for both my desktop and laptop, and there was a big difference between the two.

Desktop - 10min, 46 seconds Matching Time, 43 minutes, 6 seconds alignment time
Laptop - 17min, 22 seconds matching time, 14 minutes, 28 seconds alignment time

Same processing parameters of 136 images, high accuracy with reference preselection, key point limit 40,000 and tie point limit of 15,000.

Yes, hyperthreading is enabled on both machines and utilizing all cores/threads to 100 percent.


General / Re: Slow alignment time - compare between computers
« on: September 21, 2016, 04:56:59 PM »
Thanks for the comments guys, but I have done a number of checks on my desktop and all things seem to spec out fine on it. I ran a benchmarking program I found, and the dual CPU benchmarks over twice the rating of my laptop, so it seems to be fine as well. I also had my IT guys swap out my old RAM (64GB), but it operates the same as my 128GB.

I was wondering about Anti-Virus, etc., but it is weird to me that the Alignment step is the only process that is slower than my laptop, and just the one step of the Alignment Time. In looking at the CPU usage during this step it is running 100 percent of all cores as well.

Just to check, I also just reran an old project from a year ago that was done on the same desktop, and I am now seeing that same much slower alignment time as well. See the attached screenshot. The matching time is virtually identical, but the old "Alignment Time" was 3 minutes and 7 seconds, while the new was 26 minutes 21 seconds.

I am not sure if I just have some incompatibility on my desktop right now, but still see it as odd that it is only that one step that it is having a serious issue with.


General / Slow alignment time - compare between computers
« on: September 14, 2016, 06:16:10 PM »
I have an issue with our main processing workstation right now, with much slower performance of it to a laptop of mine, but only for certain processes in Agisoft. The attached image outlines the two areas, with the Alignment time being much slower on my desktop than my laptop.

The desktop is a dual Xeon E5-2650 V2 with a Quadro K5200 and 128GB of RAM, whereas my laptop has in I7-3740QM with a Quadro K4000M with 32GB of RAM. Both running Windows 7 Enterprise.

Every other process is much faster on my desktop, except for the Alignment Time. The matching time is faster, but not the alignment.

I am not sure why this would be so much different, and I have also played with all the OpenCL settings as well.

Any ideas would be appreciated.


Pages: [1] 2 3