Matrice 4TD DJI Pilot Thermal Mission blurry IR's

[Tlaw]

All,

Has anybody seen this before?....my RGB's are clear but my IR's blurry....first time this has happened and I cant think of anything ive changed....

 

[Tlaw]

Follow up: I rebooted and it resolved. Sending log/issue feedback to DJI...will advise if I get answers

 

[jaja6009]

It looks like the thermals are zoomed in?

How are your thermal orthos turning out?
I have yet to find a workflow that works every time.

 

[Tlaw]

It looks like the thermals are zoomed in?

How are your thermal orthos turning out?
I have yet to find a workflow that works every time.

ZERO issues on dozens of automated DJI Pilot 2 area missions for solar jobs previous with this M4TD......sent images & logs to DJI...will advise...and agreed it looks like they are zoomed

 

[Mflyer]

Possible thermal assignment coming up here.

Not to hijack the thread but what software do you prefer for the ortho map and analysis reporting?
I'm shopping around but the ready to use software is awfully expensive ..

 

[jaja6009]

Possible thermal assignment coming up here.

Not to hijack the thread but what software do you prefer for the ortho map and analysis reporting?
I'm shopping around but the ready to use software is awfully expensive ..

DJI TAT (Thermal Analysis Tool) makes good, simple reports if you have to do quantitative analysis.

If you want eye candy, you can make an orthomap. The map can be radiometric (having temperature values) or nonradiometric.

If you want radiometric I only know of three main stream solution with traditional photogrammetry, Agisoft Pro, Pix4Dmapper and Drone Deploy. If you go with Agisoft and Pix4D, you will need to convert the proprietary DJI RJPGs to grayscale TIFF. There is a free app out there from a guy named Miro Rava that works great. He has even updated it for the Matrice 4T.

I recently ran a dataset where I flew nadir, single grid with the aircraft locked facing the same orientation as the solar panels. It came out good in both Agisoft and Pix4Dmapper, but Pix4D was the better output.

Pix4Dmapper, Index Map. Every pixel contains a temperature value. You can change thermal parameters in the TIFF conversion software such as emissivity, TRFlec and some others.




The temperature values from the index map and DJI Thermal Analysis Tool were very similar for quantitative analysis.

Pix4Dmapper got maximum 136 degrees F for this anomaly.

 

[Mflyer]

All for eye candy indeed. From a user perspective, the clickable Ortho showing AI driven anomaly and GPS location seem to win ..
I like those solutions a lot myself. Looks like Sitemark solar software is one of them but very expensive.

Pix4D looks usable however. I see it also has a monthly subscription which is not that steep in price. Still would need to validate found anomalies and create a report for it. Also would the GPS data still be present after converting to TIFF?

It's funny, we like buying expensive cars, but don't wanna pay for good tires : ]
Thanks for your knowledge jaja6009

 

[jaja6009]

Yes, the image GPS metadata is still present.

 

[Tlaw]

DJI TAT (Thermal Analysis Tool) makes good, simple reports if you have to do quantitative analysis.

If you want eye candy, you can make an orthomap. The map can be radiometric (having temperature values) or nonradiometric.

If you want radiometric I only know of three main stream solution with traditional photogrammetry, Agisoft Pro, Pix4Dmapper and Drone Deploy. If you go with Agisoft and Pix4D, you will need to convert the proprietary DJI RJPGs to grayscale TIFF. There is a free app out there from a guy named Miro Rava that works great. He has even updated it for the Matrice 4T.

I recently ran a dataset where I flew nadir, single grid with the aircraft locked facing the same orientation as the solar panels. It came out good in both Agisoft and Pix4Dmapper, but Pix4D was the better output.

Pix4Dmapper, Index Map. Every pixel contains a temperature value. You can change thermal parameters in the TIFF conversion software such as emissivity, TRFlec and some others.

View attachment 31319


The temperature values from the index map and DJI Thermal Analysis Tool were very similar for quantitative analysis.

Pix4Dmapper got maximum 136 degrees F for this anomaly.

Extra crispy plz