AUTEL EVO II Pro 6k camera sensor

[ñako]

Hello.

In order to do some Photogrammetry, I need to know the size of the sensor (in mm) of the 6k camera in my EVO II Pro to calculate the Ground Sample Distance of my orthomosaic. I have seen already that it is said in the app missions but I want to be sure.

I have found that EVO II Pro camera builds a SONY IMX383.
Please, could you confirm me that is this the sensor built in the 6k camera, please?

If Sony IMX383 is the sensor, there is something that I do not understand when checking the specifications within the SONY product information sheet (attached).

It says that the pixel size = 2,4 µm.
With a maximal sensor size of 5472 X 3648 pixels, it only reaches a dimension of 13,1328 mm x 8,7552 mm. These dimensions are far from the 1" size it is suppoused to be, isn´t it? An inch is equal to 25,4 mm if I am not wrong.

I expected to find a 36x24 or 36x25,4 mm full frame sensor.

Could you explain what am I doing wrong, please?

Greetings,
Ignacio

 

[Brock Inspector]

You're posting this as a quasi-engineering question, but it is a question that has been asked and answered countless times. Here, from19 years ago, for instance:

Making (some) sense out of sensor sizes

We regularly receive emails from readers confused as to the actual size of sensor used in digital cameras. Sensors (CCD / CMOS) are often refered to with an imperial fraction designation such as 1/1.8" or 2/3", this measurement actually originates back in the 1950's and the time of Vidicon...

www.dpreview.com

The 1" Vidicon tube naming convention is not a problem unique to Autel Evo II Pro

 

[Tom skerke]

Awesome

 

[Claudio NC]

The good and old Ken Rockwell has also for this topic the right words to comment: One-Inch Sensors

 

[agrofactory]

Hello.

In order to do some Photogrammetry, I need to know the size of the sensor (in mm) of the 6k camera in my EVO II Pro to calculate the Ground Sample Distance of my orthomosaic. I have seen already that it is said in the app missions but I want to be sure.

I have found that EVO II Pro camera builds a SONY IMX383.
Please, could you confirm me that is this the sensor built in the 6k camera, please?

If Sony IMX383 is the sensor, there is something that I do not understand when checking the specifications within the SONY product information sheet (attached).

It says that the pixel size = 2,4 µm.
With a maximal sensor size of 5472 X 3648 pixels, it only reaches a dimension of 13,1328 mm x 8,7552 mm. These dimensions are far from the 1" size it is suppoused to be, isn´t it? An inch is equal to 25,4 mm if I am not wrong.

I expected to find a 36x24 or 36x25,4 mm full frame sensor.

Could you explain what am I doing wrong, please?

Greetings,
Ignacio

Hello Ignacio,

How you were able to know the specific sensor that the EVO II Pro mounts? I'm having a hard time trying to figure out the true focal lenght of the sensor (not the 35 mm equivalent) in order to calculate properly GSD. I know Autoexplorer automatically gives you the GSD but I would like to know the fundaments involved in it.

Thanks

 

[ñako]

Hello Ignacio,

How you were able to know the specific sensor that the EVO II Pro mounts? I'm having a hard time trying to figure out the true focal lenght of the sensor (not the 35 mm equivalent) in order to calculate properly GSD. I know Autoexplorer automatically gives you the GSD but I would like to know the fundaments involved in it.

Thanks

Hi,

You can find it with a small research in internet like the Amazon.com website Amazon.com
And get the specs in pdf in SONY website: https://www.sony-semicon.co.jp/products/common/pdf/IMX383-AAQK_Flyer.pdf

If you could find out why the GSD in Autel explorer app is different form the one you get with the formula, just let me know please.

Greetings

 

[Saladshooter]

How you were able to know the specific sensor that the EVO II Pro mounts?

It was widely reported when the EVO 2 Pro was released that the sensor was the Sony IMX383. The info originated from Autel to express the quality of the camera. I have not seen the true focal length of the camera posted anywhere.

 

[Brock Inspector]

Two brand new members discussing the most arcane discrepancies sounds line one troll talking to their many selves. Who is Ignacio?

 

[Saladshooter]

Two brand new members discussing the most arcane discrepancies sounds line one troll talking to their many selves. Who is Ignacio?

Ignacio is the the original poster from last March. He signed his post.

All I see in this thread is a couple of people trying to get some information so that they can use their EVO 2 Pro in photogrammetry applications........ trolls usually jump right into the complaints.

The actual focal length is important for figuring out overlap between pictures.... I imagine Autel should have this info available, if you can reach the right person. The EVO 2 should be good for this type of work though it is an electronic shutter. The long flight times are pretty advantageous.

 

[agrofactory]

Two brand new members discussing the most arcane discrepancies sounds line one troll talking to their many selves. Who is Ignacio?

I was just trying to fully understand how Autel explorer app calculate GSD value since there's no official info about true focal lenght of the sensor on the website, documents, etc. Maybe I'm naive expecting to get that kind of specs from the manufacter itself... or maybe just look on the wrong place.

I don't get why that is a troll thing.

It was widely reported when the EVO 2 Pro was released that the sensor was the Sony IMX383. The info originated from Autel to express the quality of the camera. I have not seen the true focal length of the camera posted anywhere.

Well, the only info about the sensor that I could get it's from online stores. Again, just trying to get some official info about it.

 

[dhunter]

I agree with Ignacio, the original author of this thread.

The availability of the actual camera information you need to calculate the GSD for an EVO II Enterprise drone is very difficult. I believe in trust and verify. As important as GSD is to accuracy and quality of photogrammetry work outputs, I have trouble trusting a number calculated by the software in the background. Autel has marketed too many different models of the EVO II product line so it is hard to tell which camera specifications apply for the non-thermal EVO II Enterprise V2 which I purchased and am waiting on delivery this weekend.. The EVO II Pro camera appears to be slightly different than the EVO II Enterprise V2 model.

I have found the following in doing a Google Search for the EVO II Pro 6K camera (Pix4D):

Pix4D Supported drones
https://support.pix4d.com/hc/en-us/articles/360045923031-Supported-drones-and-cameras-PIX4Dreact

SXT705_10.6_5472x3648 (RGB)???

This does not appear to match the camera specs listed on Autel's websites. The solution is probably easy. If a forum member with an EVO II Enterprise V2 model would post the portion of a quality report from whatever photogrammetry software they use, the camera data should be listed to determine the solutions to the formulas below:

GSD = Sensor Width * Working Distance / True Focal Length * Image Width
Crop Factor = 35mm / Digital Sensor Width
True Focal Length = 35mm Equivalent Focal Length / Crop Factor
Motion Blur = Camera Shutter Interval (s) x Drone Speed (m/s)
Optimal Flight Speed = Image Interval / Camera Interval
Sensor Width = 2*(Focal Length*TAN((0.5*Horz Field of View)/57.296))

Thanks in advance and I look forward to some feedback from the more experienced forum members.

 

[ikebenaka]

Did you ever find the true focal length/sensor width measurements? I need them for my photogrammetry work and I'm finding conflicting information as well...

 

[mikear]

To revive this thread I'd like to share my experience and calculations. Talking about the EVO II Pro V3, Autel tells there is a 1" type of sensor which has a 13.2/8.8mm real size (sony specs of IMX383). According to my calculations (see screenshot) this would give a real focal length of 10.633mm, which is more or less in line with the info given on the Pix4D site (SXT705_10.6_5472x3648 (RGB)).
What puzzles me is that when I take a picture with the drone exactly 1m from the wall and I measure the visual width on the wall, I do not find a ground image width of 1.241m (as calculated in the image below) but it is more close to 1.38m. Which would lead to 26mm (35mm)equivalent focal length and a 9.53mm absolute focal length.
This would mean that the specs by Autel are not correct, or my calculations incorrect.

 

[ikebenaka]

Hey Mikear,

I can't believe this thread was revived! Thanks for sharing your experiment results. I have been trying to work out the same issue for my EVO II Pro V2 because I am trying to conduct photogrammetry on whale morphology. It has the same optical specs as the V3 based on your table.

Here is my take: the focal length is not changing, and shouldn't because it's based on the physical dimensions of the EVO's lens. What might actually be happening is that the drone is applying a distortion correction for barrel distortion, and then crops out the non-rectangular borders of the image to maintain a 4:3 aspect ratio. The cropped image is then upscaled back to the original resolution of 5472 x 3648. This means you’ll underestimate the true GSD, because the effective image width in meters (after cropping) is smaller than the raw, full-sensor FOV implies. I believe that this would result in an overestimation of real-world length in meters for your wall.

Check out the attached paper by Segre. It explains an experiment that could help you find the true GSD.

I plan to carry out this experiment next week and can share my results.

Please write back with any questions or rebuttals to my take. Figuring this stuff out with multiple people is a lot easier than doing it yourself as I've found!

 

[mikear]

Hi Ikebenaka, thanks for sharing this recent article (for those who would like to read the article online: https://onlinelibrary.wiley.com/doi/10.1111/mms.70022), I think it is highly valuable for those who want to do photogrammetry with consumer level drones like the Autel EVO II's. I'll need some studying on this article but definitely it will increase my knowledge and maybe of others, and the quality of measurements if not RTK is available (and maybe even then...).
I'll surely check if in the settings of my RC device any digital correction of distortion is turned on and will come back with my findings.
Thanks again for this reference and your thoughts, extemely helpful!