Welcome to our Community
Wanting to join the rest of our members? Feel free to sign up today.
Sign up

Anyone ever used a Parrot Sequoia?

consultant

New member
Joined
May 25, 2019
Messages
4
Reaction score
0
Testing a Parrot Sequoia and the RGB photos are terrible. It's not a drone speed issue as I tried even on the lowest speed setting in Pix4D Capture. When you have motion blur issues it's consistent throughout the image. This is the strangest thing I've ever seen in that it has what looks like bands of blurriness as some bands of areas like more detailed than others. Trying to determine if I have a bad RGB camera sensor. The other camera the images are consistent detail throughout the frame.

This is 4 inch/pixel GSD (135 feet AGL)


I know 4 in/pixel is not very high resolution but the images look very abnormal. Again, this is flying at slowest speed setting. There was absolutely no difference in the image between the fastest and slowest setting.
 
Just excessive vibration interacting with the camera's rolling shutter.
 
That would be pretty bad then considering the photos were taken with a Parrot Bluegrass which was made for having the sequoia built-in. Are you telling me the drone specifically engineered to carry a Sequoia has vibration issues making the RGB sensor essentially unusable?
 
Not necessarily, you could have a bad/imbalanced prop or a mounting issue.
 
How do you verify if that's the case? Try different props, or would the drone hovering close up have vibration to the point it you could see the drone slightly shaking? It appeared to hover without any noticeable shaking. And it doesn't seem logical to me that all areas of the photo frame have about the same blur.
 
Testing a Parrot Sequoia and the RGB photos are terrible. It's not a drone speed issue as I tried even on the lowest speed setting in Pix4D Capture. When you have motion blur issues it's consistent throughout the image. This is the strangest thing I've ever seen in that it has what looks like bands of blurriness as some bands of areas like more detailed than others. Trying to determine if I have a bad RGB camera sensor. The other camera the images are consistent detail throughout the frame.

This is 4 inch/pixel GSD (135 feet AGL)


I know 4 in/pixel is not very high resolution but the images look very abnormal. Again, this is flying at slowest speed setting. There was absolutely no difference in the image between the fastest and slowest setting.

Without knowing anything about your camera, I can observe your picture taken just before the sun goes down (long shadows), which in turn means your camera must be running at a long shutter speed. Try to take a picture around midday instead and see it doesn't help.

I can also see you have shakes in the picture from one or more props out of balance. I would recommend you buy new 2 sets and use them instead. Balancing a three-blade propeller is not so easy.

Good luck, Leif.
 
Vibration would be tiny and high frequency, so invisible. Best is to try other props.

which in turn means your camera must be running at a long shutter speed
He said he already tried different shutter speeds manually without a difference. And that's correct, vibration-induced blur won't be affected.
 
I have a second brand new set of props as backups and they make no difference. If there was a vibration issue, the other cameras on this multi-spectral camera would be blurry too.

Notice how the title of this thread says "Anyone USED a Parrot Sequoia?" If you haven't used one than your just shooting in the dark (no pun intended). But I still appreciate the suggestions.

The same banded blurriness was seen from photos I took on a bright sunny day.

Someone in another forum that actually has used one of these cameras has confirmed this is normal due to the camera's small/slow sensor and the rolling shutter is what creates the banded blur effect. The other sensors have a global shutter which explains why they look consistent.

So at least now I know there isn't a problem with the drone or the camera or I'm doing something wrong. I did notice the photos were slightly better but very negligible flying at one speed setting lower from the fastest in Pix4D.

Essentially, the bottom line is this is what you can expect from a puny 1.2MP sensor with rolling shutter, at least on this camera. People thinking it was a vibration issue or too low of light just have never actually used this camera.
 
It is due to the vibration interacting with the rolling shutter as I said right at the beginning, that is the correct technical explanation.
If the aircraft didn't vibrate that sensor would not give distorted images, if that sensor didn't have rolling shutter then the aircraft could vibrate without causing that issue.

The package as it is delivered happens to give that result then. But regardless, it still stands that if you could find a way to reduce vibration that gets to the camera you could mitigate the issue and get better results.
 
Last edited:
  • Like
Reactions: PoliYa
I have a second brand new set of props as backups and they make no difference. If there was a vibration issue, the other cameras on this multi-spectral camera would be blurry too.

Notice how the title of this thread says "Anyone USED a Parrot Sequoia?" If you haven't used one than your just shooting in the dark (no pun intended). But I still appreciate the suggestions.

The same banded blurriness was seen from photos I took on a bright sunny day.

Someone in another forum that actually has used one of these cameras has confirmed this is normal due to the camera's small/slow sensor and the rolling shutter is what creates the banded blur effect. The other sensors have a global shutter which explains why they look consistent.

So at least now I know there isn't a problem with the drone or the camera or I'm doing something wrong. I did notice the photos were slightly better but very negligible flying at one speed setting lower from the fastest in Pix4D.

Essentially, the bottom line is this is what you can expect from a puny 1.2MP sensor with rolling shutter, at least on this camera. People thinking it was a vibration issue or too low of light just have never actually used this camera.

I don't think you should expect a nice picture from that camera as it uses several sensors with different resolutions that are probably added together. The spectrum you see is from the blue color up to the infrared area (NIR) on a single image. What you use is not sharpness but the reflection of the light from the crops to see if there is disease or deficiency diseases. When you mix the infrared area into an ordinary image, it looks very woolly and the colors are wrong. I have a ZWO color camera that can cover up in the IR-A area which I sometimes use on my telescope without an IR cut filter. Sometimes I use it only as an IR camera and that looks strange.

However, the IR area does not explain the strange shakes that you have on a part of the picture. On reflection, can the "shaking" be a product of the image processing that the camera does ?


Regards, Leif.
 
I don't think you should expect a nice picture from that camera as it uses several sensors with different resolutions that are probably added together. The spectrum you see is from the blue color up to the infrared area (NIR) on a single image. What you use is not sharpness but the reflection of the light from the crops to see if there is disease or deficiency diseases. When you mix the infrared area into an ordinary image, it looks very woolly and the colors are wrong. I have a ZWO color camera that can cover up in the IR-A area which I sometimes use on my telescope without an IR cut filter. Sometimes I use it only as an IR camera and that looks strange.

However, the IR area does not explain the strange shakes that you have on a part of the picture. On reflection, can the "shaking" be a product of the image processing that the camera does ?


Regards, Leif.
I believe that the Sequoia camera does not take into account the blue bands directly. "The ordinary image" is captured by a specific RGB sensor wich is a rolling shuter (unlike multispectral sensors that are in global Shuter)...
This is probably the reason for the low quality of this ordinary image (that of "consultant"). I think the orthomosaic created by the Sequoia camera (by its dedicated RGB) is not the priority of this sensor but rather what it offers in terms of "invisible" analyses (more useful to farmers). Micasense sensors (another subsidiary of parrot in agricultural field) are more suitable for a quality orthomosaic but with a higher price...A24B5C67-5E91-4727-BECA-C88FABB711F2.jpeg
 
I have a second brand new set of props as backups and they make no difference. If there was a vibration issue, the other cameras on this multi-spectral camera would be blurry too.

Notice how the title of this thread says "Anyone USED a Parrot Sequoia?" If you haven't used one than your just shooting in the dark (no pun intended). But I still appreciate the suggestions.

The same banded blurriness was seen from photos I took on a bright sunny day.

Someone in another forum that actually has used one of these cameras has confirmed this is normal due to the camera's small/slow sensor and the rolling shutter is what creates the banded blur effect. The other sensors have a global shutter which explains why they look consistent.

So at least now I know there isn't a problem with the drone or the camera or I'm doing something wrong. I did notice the photos were slightly better but very negligible flying at one speed setting lower from the fastest in Pix4D.

Essentially, the bottom line is this is what you can expect from a puny 1.2MP sensor with rolling shutter, at least on this camera. People thinking it was a vibration issue or too low of light just have never actually used this camera.
The picture seems odd, indeed.
But I believe it depends if this image is viewed at the exit from the camera or through the algorithm of a software like Pix4D Fields (even Pix4D Mapper).
As Parrot is the owner of BlueGrass drone and Pix4D, the Sequoia camera is particularly suitable for the software that "reworks" the images taken in flight for further analysis.
Do you read this image like any photo or via Pix4D Fields ?
 
Last edited:
I believe that the Sequoia camera does not take into account the blue bands directly. "The ordinary image" is captured by a specific RGB sensor wich is a rolling shuter (unlike multispectral sensors that are in global Shuter)...
This is probably the reason for the low quality of this ordinary image (that of "consultant"). I think the orthomosaic created by the Sequoia camera (by its dedicated RGB) is not the priority of this sensor but rather what it offers in terms of "invisible" analyses (more useful to farmers). Micasense sensors (another subsidiary of parrot in agricultural field) are more suitable for a quality orthomosaic but with a higher price...View attachment 2132

Yes ok. Parrot then use a single camera for each of the 4 "colors" the camera covers. Yes the blue spectra did not appear to be covered. I'm not aware of how it's built up and what colorspace it covers. My knowledge is only how one could have made out from the few data I have seen.

Yes the camera does not need to be able to set completely sharp but must reproduce the amount/number of photons correctly.

Regards, Leif.
 
Hello,
I wish I had read this thread before buying 3 sequoia... The multispectral images are fine, but the RGB are piece of crap!! The advertised results are absolutely lies, and I'm really upset about the parrot sales team that promote this camera. This is not a rolling shutter issue, I would say it is a quality of the sensor/lens issue.
The specs says 4.1cm of resolution at 150m... look at the attached image, you'll clearly see that it's a joke... I tried with different sequoias on different UAVs, same blurry / non focused images. They are so bad that even with 80/80 overlap, Pix4D can't calibrate them...

When you call parrot for information or help, you only have a call center from india that says you can check for parts or buy another one on Amazon!?! Incredible...

I have several sensefly Ebees, that are equipped with SODA or Sequoia, and I have now to re-capture every project with SODA for the RGB maps, that mean a lost of thousand euros for me...
Conclusion is, if you need only multispec survey, sequoia is OK, but if you need RGB too, go for a micasense.

Lasts informations:
-The download of the data is totally unreliable. I spent hours for getting my pictures out of the cameras, with corrupted files an interrupted operations.
-The use of the SD card is prohibited as the camera is getting extremely hot, stop to write on it and stop to be recognized by the ebee. I only use the internal memory because of that. So when you are in the field with a full memory and need to empty it, you loose your day... :(
 

Attachments

  • IMG_200709_125704_0026_RGB.JPG
    IMG_200709_125704_0026_RGB.JPG
    1.8 MB · Views: 8
  • IMG_200709_113938_0398_RGB.JPG
    IMG_200709_113938_0398_RGB.JPG
    1.6 MB · Views: 8

New Posts

Members online

Forum statistics

Threads
5,294
Messages
45,048
Members
7,964
Latest member
edamame