Inconsistency flight speed


#1

I’m experiencing inconsistent gaps between images as a result of increased speed.

Using a DJi Phantom 4 drone
my overlaps are Front 75 and side 65 .
what am I doing wrong?


#2

It’s not the flight speeds, it’s the shutter timing. This has been happening inconsistently for a while, but seems to happen to me more on larger jobs with long runs.

@Anya @Jamespipe


#3

So, what’s the solution?


#4

I wish I had a way to measure the distances in between each shot. This would tell us if the system is experiencing a minor lag until it eventually misses the timing and then double-shoots. That’s just my guess though. Dronedeploy is well aware, but we have to keep firing at it. Please use the other thread for all future comments.


#5

Plot your geotags in GIS and measure and script a “distance from previous point” model. I’ve read some of your other posts, so I’m assuming you’ve already thought of this. If so, why doesn’t that solve your problem?


#6

That’s a good idea, but I am not very experienced in QGIS. I’m assuming there is an easy way to plot all the geotags for 100’s of points and then report the inverses between each point? I can do this in CAD, but don’t know where to start in QGIS.

I have gone through several more jobs and it is too inconsistent to be what I am thinking. I am doing some testing on a 3-4 acre stockpile and it too had some inconsistent shot spacing and there were only about 50-60 images. I cycle my mSD cards so that’s not it. DroneDeploy is well aware and we are starting on the 2019 goals so I’m sure this will come up again.


#7

Unfortunately, I don’t use QGIS either. All my experience is in ESRI platforms. I typically use an EXIF extractor to create a table with the point locations.
Here’s a workflow: Plot the points in GIS, convert points to lines (based on photo order (onsecutive ID)), then attribute distance for each line, spatial join the line attribute table back to the point table (based on the point at the line end). You’ll then have a table with all info needed. You could then, based on the time stamps, calculate the speed travelled between each photo point Then you would know if it’s possibly an airspeed issue. The rusulting table would have all the other exif metadata in it as well, so you could look at the speed of the aircraft vs. shutter speed, etc.
You can create a data processing model that will just do all this for you and spit out a table for every dataset.


#8

Although the waypoints are uploaded to the aircraft before flight, the photos are triggered each time from the tablet signalling through the RC. I haven’t confirmed it, but I think the issue may be the tablet not sending the trigger signal at the proper time due to the processor getting backed up.

I see the spacing get worse on flights where my iPad 4 Mini starts heating up, which causes the OS to start slowing the processor down to reduce heat.


#9

@RWR, That make sense, thanks for the breakdown.

@HawkViewAerial, Thanks for the additional thoughts. Definitely something to take into consideration. I use a 5th gen iPad.

@timmyd, what kind of device are you using?


#10

I use the ipad Mini 4 and I have been having problems with the spacing and grouped up photos. I primarily use Pix4D for flight planning. Too bad we can not specify the P4P to just take a photo every x seconds. Then you could have a chart to adjust your speed for the desired overlap.

In Pix4D if you select “Fast Mode” the drone will continue to take pictures even if the connection is lost with the remote, so that seems to rule out that it is a processor problem on ipad mini side?? Hopefully someone will find out what the problem is.


#11

Michael,
I imported the Post Processed events file for 146 photos that I did on the video using this guide.
https://www.qgistutorials.com/en/docs/importing_spreadsheets_csv.html

I used the distance matrix using this guide:

Not sure if that helps any or not.


#12

Meant to show full screenshot with imported points.