Perseverance Imagery, technical discussion of processing, cameras, etc. Options

[Ant103]

As for me, I'm using ImageJ with Debayer plugin. Alongside with a batch script, I can do hundred of pictures in a very short time.

[James Sorenson]

I'm using Pixinsight which is an astrophotography processing suite, it has a real nice debayer script. I use it for my astrophotography and it is expensive, 300 USD, but for those that can't afford that, the Planetary Image Pre-processor (PIPP) has a great batch debayer function in it which I've also used and it works great. On the plus side it's a free and easy program.

[phase4]

The Marslife website uses an open source GPU shader to debayer images 'on-the-fly' which is blazingly fast.
While not really useful for non-coders I thought to share this solution as a reference. Pretty well documented, too.

Efficient, high-quality Bayer demosaic filtering on GPUs by Morgan McGuire.

[fredk]

The Marslife website uses an open source GPU shader to debayer images 'on-the-fly' which is blazingly fast.

Thanks; I was wondering how you deBayered the mastcam sequences so fast.

Incidentally the Malvar etal interpolation scheme you use is the same as that used onboard MSL to deBayer when jpegs are sent down; see the MSSS camera specification. (And that's the scheme I use in my own code.)

Anyway, deBayering M20 frames is a breeze - dealing well with the jpeged Bayered images for MSL is hard work!

[htspace]

i use Gmic's built in program " bayer2rgb"
https://gmic.eu/
it is a commandline tool

CODE

gmic InputImage.png bayer2rgb 1,1,1 -o OutputImage.png

now they will be a bit "yellow/green" do to the contrast stretch that these png's have had done to them

Thank you！

[htspace]

I must admit - I've found Gmic's debayer performance to be pretty ugly - especially at the edge of images.

I've been using PIPP which can batch debayer images and I've been very pleased with the results

https://sites.google.com/site/astropipp/

I don't know how to perform batch operations, only one picture can be output.can you teach me? Thank you!

[htspace]

I'm using Pixinsight which is an astrophotography processing suite, it has a real nice debayer script. I use it for my astrophotography and it is expensive, 300 USD, but for those that can't afford that, the Planetary Image Pre-processor (PIPP) has a great batch debayer function in it which I've also used and it works great. On the plus side it's a free and easy program.

Yes, this software is expensive.
I don't know how to perform batch operations, only one picture can be output.can you teach me? Thank you!

[James Sorenson]

For PIPP, all you need to do is open up all of the images that you are going to debayer, then on the input tab select the option "Debayer Monochrome frames" and the pattern to RGGB. You can also choose the processing algorithm as well if you want VNG or Bilinear works good, you can experiment. To see a preview before final processing, you can click the "Test Options" button in the top right. If it looks good, you can move to the output tab and choose your output file options. Once ready, just go to the "Do Processing" tab and click "Start Processing". All the images that you loaded should be batch debayerd into a new folder.

[Greenish]

If "camera_position" is correct at (2.08195,-0.111422,-1.69411) then WATSON's eyeball would be a fairly high 1.7m off the ground, ...
... Not convinced ...

Well, maybe I'm not totally nuts. These pictures are before & after the WATSON test images. Looks pretty high to me.

Implies roughly the scale bar shown below, so the larger rocks are ~2cm.

[htspace]

For PIPP, all you need to do is open up all of the images that you are going to debayer, then on the input tab select the option "Debayer Monochrome frames" and the pattern to RGGB. You can also choose the processing algorithm as well if you want VNG or Bilinear works good, you can experiment. To see a preview before final processing, you can click the "Test Options" button in the top right. If it looks good, you can move to the output tab and choose your output file options. Once ready, just go to the "Do Processing" tab and click "Start Processing". All the images that you loaded should be batch debayerd into a new folder.

Thank you! The previous steps were successful, and finally failed.I don't know how to solve it.

[htspace]

As for me, I'm using ImageJ with Debayer plugin. Alongside with a batch script, I can do hundred of pictures in a very short time.

Oh it's you! Thank you for your work!
How to create a batch script？I am a novice……

[htspace]

For PIPP, all you need to do is open up all of the images that you are going to debayer, then on the input tab select the option "Debayer Monochrome frames" and the pattern to RGGB. You can also choose the processing algorithm as well if you want VNG or Bilinear works good, you can experiment. To see a preview before final processing, you can click the "Test Options" button in the top right. If it looks good, you can move to the output tab and choose your output file options. Once ready, just go to the "Do Processing" tab and click "Start Processing". All the images that you loaded should be batch debayerd into a new folder.

I found the reason, it is the system language problem.

[Art Martin]

I just installed PIPP and it seems to be a very full featured program but I'm certainly lacking a great deal of info. I downloaded a Watson image of the underbelly that to my untrained eye seems like it's a candidate for debayering since there are lines and dots across the image that makes it look blurry and unprocessed. However, when processing it through PIPP it didn't seem like it did anything. The output looked pretty much like the input and it certainly was not in color. There were tons of options to choose but unfortunately the examples on the website were for some Jupiter video file not some single image. Any shortcuts someone can point to?

Edit: I did manage to get a color image debayered using Gmic by just trial and error. I'll post it in the helicopter area.

[mcaplinger]

If "camera_position" is correct at (2.08195,-0.111422,-1.69411) then WATSON's eyeball would be a fairly high 1.7m off the ground...

I take it back, perhaps this is correct. Should be possible to find the WATSON footprint in the front Hazcam image.

I've asked if the focus position can be put back into the public data, but we don't have much control over that.

[fredk]

Has anyone figured out how to sort the images at mars.nasa.gov/mars2020/multimedia/raw-images/ by time taken? Selecting a sort in the "sort by" box gives images out of order according to the listed LMST. (Eg, the first few images are out of order when you filter on sol 18.) The corresponding operation on the MSL site works fine.

The corresponding api call

CODE

https://mars.nasa.gov/rss/api/?feed=raw_images&category=mars2020&feedtype=json&num=50&page=0&order=sol+desc&&condition_2=18:sol:gte&condition_3=18:sol:lte&extended=sample_type::full,

similarly gives an out of order json file, as do any alternatives I've tried for the "order=" option.

Maybe this is related to the timestamp problems that are being worked on?