Topics

1

General / Illogical GSD vs overlap results

« on: September 27, 2021, 09:48:12 AM »

Hi, fellow metashapers

A collegue of mine did some flights for his master thesis and ended up with some apparently illogical GSD results. I have tried to find some logic into it but couldn't.

He combined different tilt angles with overlap percentatges. The results are shown in the table image below.

Can somebody explain how a 30% tilted angle with less overlap is yelding better GSD than with 90% overlap? Does this have a scientific explanation or is it a software misscalculation?

Thank you very much and have a good week.

2

General / Merging Chunks for Multispectral Datasets

« on: June 21, 2021, 12:06:11 PM »

Hello all,

I am working with a quite large study area where I have to perform several drone flights in order to cover everything with a multispectral camera.

Before each take-off I take a picture of a reflectance panel to later correct each flight radiometry.

I process each flight in a different chunk.

My doubt here is, if I merge the chunks, is the radiometric calibration kept for each batch?

If I were to process everything at once, can i group the cameras and apply different radiometric calibrations to different camera groups in the same chunk?

Thank you very much!

3

Bug Reports / Bug in reflectance adquisition

« on: November 12, 2020, 04:42:22 PM »

Hello, dear community,

I have been recently flying for a study over a forest plot with a Parrot Sequoia. We had to fly the plot in two batteries, so we had to reprogram the Sequoia again for the second flight.

When reconstructing the multi images we found this:

https://ibb.co/wMG8P3N

It looks like a malfunction of the sun sensor, apparently.

Has this happened to anybody? Does someone know what can it be and, more important, if it can be fixed, somehow?

Thank you in advance! 

4

Bug Reports / Weird Orthomosaic reflectance

« on: September 24, 2020, 06:05:44 PM »

Hello everyone,

I amb having a problem with a project I am working with. We are trying to map a forest plot where there has been different post-fire management operations. To do it, we used a DJI Phantom 3 Adv with a Parrot Sequoia multispectral camera installed.

We designed the flight to map the entire plot at once. We first took the calibration data using a reflectance panel from AIRINOV. Then, we set up the Sequoia (through it's beloved web interface) and then uploaded the flight plan using Map Pilot. It is kind of a big plot, so the dron returned when it had mapped about 50% of the entire surface. As the Sequoia drinks from the drone battery, we first changed the battery, set up the Sequoia again, re-upload the flight plan using Map Pilot, and it automagically resumed the flight from the last known point. So far, so good. It finished the plot with the 2nd battery.

The problem appeared when processing all those images. As you can see in the orthoimage, something happened between the first and the second part of the flight. One part has a lot more brightness or reflectance than the other. The raw images are okay, they are not "burned" so I have hope that something can still be done.

Can anyone help me on this?

Thanks in advance

5

General / GCPs in Multispectral chunks

« on: July 18, 2018, 08:48:29 PM »

Hello, Community experts,

I am working with multispectral imagery for the first time.

When I was manually pinning the GCPs in the pics I realized that they were only being set in the band I had as a 'Primary Channel' in that precise moment. If I change the band, the markers have to be replaced. It makes sense because there are 5 sets of images from 5 different bands from 5 different cameras (Rededge MicaSense).

My question is then: Should I put the markers 5 times? One for each band to then be able to export precise georeferenced orthomosaics for each band?

I hope I made myself clear as it is somehow difficult to explain. If not, tell me and I will rephrase it.

Thanks!

6

General / Depth Filter Advice

« on: May 04, 2018, 09:07:48 PM »

Hi community,

I've got this project where we try to model vegetation to get height as much accurate as we can. The pics are just 8MPx each but there is more than 90% overlap. However there is a lot of detail. You can see pics attached. So:

- I managed to get 1073 pics aligned with over 1,3M Tie Points.
- I run the Dense Cloud Point in Medium with Depth filter disabled. Also, the bounding box got moved at some point and I didn't realised and I end up having lots of points under the floor.

So, two questions arise:

- If I run the dense cloud point again with the bounding box correctly aligned to the floor, will all those underfloor points be fitted over the floor, and by so, I will get more detail?

- As you read, I run the DCP with depth filter disabled. If I set it to Mild, do you think I will lose some heigh detail? There are small branches on the top that might get cut by the filter?

So... what do you think? Thanks in advance!

7

General / Manually introducing Camera Model

« on: April 30, 2018, 10:16:36 PM »

Hi!

We have been flying with a UAV over a forest plot. After extracting the frames with Photoshop I always lose the camera info, so when I try to calibrate it, it appears as "unknown camera". Is there any way in Agisoft to manually introduce Camera model so that the software can make the proper lens correction, etc.?

Thanks a lot!

8

Bug Reports / Input Video BUG

« on: April 25, 2018, 06:55:05 PM »

Hi all,

I am trying to slice a 10min 4K Video into single frames. Apparently, that is what "inport video" function does. Problem is, when I do that, the system just crashes. It eats all the RAM, it starts paging it into SSD till it fills it all up, and then the whole system crashes. When I check it in Activity Monitor, Photoscan is using up to 80GBs of "RAM". I have tried different things like do it with a 2sec 30fps video which should give me max 60 images. But, when I restart the system and look into the destination folder, there are like 140 images (90% of them being exactly the same) which makes no sense at all.

Are you aware of this? Am I doing something wrong? Is there a minor fix for it? Is it just with my setup?

- AgiSoft Profesinoal 1.4.1 build 5925 (64 bit).
- Macbook Pro 15' (Late 2017) i7-7920HQ @3.1 GHz
- AMD Radeon Pro 560 4GB
- eGPU Box with GeForce GTX 1060 4GB VRAM

I tried using the different GPUs, it crashes anyway.

I would really appreciate some help.

Thank you very much!

9

General / External GPU (eGPU) experience with AgiSoft Photoscan

« on: April 19, 2018, 06:04:19 AM »

Hi community,

Today I finally received a Sonnet eGFX Breakaway (eGPU enclosure). After some hours of scripts running up, uncountable reboots, crashes, etc. I managed to have it all working and wanted to share some amazing preliminary results I got.

My basic setup:

-   Macbook Pro 15’ (Late 2017)
-   CPU: Intel Core i7-7920HQ (4 Cores, 8 Threads @ 3.1 GHz, TB @ 4.1 (real 3.7))
-   GPU: AMD Radeon Pro 560 4GB
-   RAM: 16 GB 2133 MHz LPDDR3

Added:
-   Sonnet’s eGFX Breakaway Box 550 with Thunderbolt 3
-   ZOTAC GeForce GTX 1060 6GB

I used sample data from Parrot®. The sample contains 356 pictures at 4608x3456 resolution. I tried the following setups:

-   AMD + CPU
-   Nvidia + CPU
-   AMD + Nvidia + CPU
-   AMD + Nvidia

And just had time to Align and built de Dense Cloud in the lowest accuracy in boths. Here are some results.

TIME IN SECONDS (LOWER IS BETTER)


The processing times between the discrete Radeon GPU and Nvidia are jaw dropping. A 253% faster! As you can see, the best setup would be the dual GPU + CPU but, the different between this and just running Nvidia+CPU or NVidia+AMD are insignificant.

As you can see, for those of you who don't want to give up on your laptop, here it's a powerfull alternative.

That being said, if you want I can try to find some time to do some other benchmarks with Agisoft Demo sets like the statue or any other in different accuracies. I will try to do it so that those considering in buying one have a little more information.

Thanks for reading and hope you found this useful!