Support for FY_2x?

[bclswl0827]

Is your feature request related to a problem? Please describe. Sanchez doesn't support add false colour to FY_2x.

Describe the solution you'd like Add support for FY_2x, including add country borders to its false colour image.

Describe alternatives you've considered Currently it's possible to add false color to FY_2x through GeoSatSignal and CLUT, but this software only supports Windows platform. (https://github.com/eswnl/FY-2x_CLUT)

Additional context None.

[nullpainter]

Sure thing. Do you have a sample image for me to use?

In the meantime, you should be able to add support by adding a section to the Satellites.json file - see here for details.

[bclswl0827]

Of course here is one from FY_2G

https://3pfy6f-my.sharepoint.com/:i:/g/personal/bclswl0827_3pfy6f_onmicrosoft_com/EcwWTl_AQiFBu-mDAwxsdzIBwN6Bo2OpH_hfodLB6Z7NYA

[nullpainter]

Thanks! Sanchez uses regular expressions against filenames to determine which satellite configuration to use for each image. Is FY-2x a standard prefix? Or is it FY-2[A-Z]? (e.g., FY-2C, FY-2G, etc.)

[bclswl0827]

Yes, FY_2x is standard prefix

[creinemann]

Quick json file addition allows it to work, though your imagery has extra black space surrounding the earth and will have to have cropping adjusted to the json as well to properly line up everything. the crop settings below are just copied from another geo satellite. You can clearly see where the continents don't line up.

{
   "DisplayName": "FY-2H",
   "FilenamePrefix": "FY-2x_1_",
   "FilenameParser": "Goesproc",
   "Invert": false,
   "Longitude": 79.5,
   "Brightness": 1.2,
   "Crop": [
     0.003166,
     0.089925,
     0.001166,
     0.001010
   ]

[creinemann]

What software are you decoding with? I noticed some use the prefix FY_2x while the file you sent has FY-2x_

[bclswl0827]

SatDump :)

[nullpainter]

We can use a regex to support both prefixes. Are the filename date formats and the image alignment the same for all?

If so, all we require is more accurate crop values, as @creinemann noted above. Either you or I can work on this. You can either measure pixels or just divide and conquer.

I can then add the final config to git and create a new release.

[bclswl0827]

Yes, they're same for all.

[sq7r0]

unfortunately there will be a problem for exact FY-2H crop values, as every capture has drift. Sat orbit is unstable which results in xy offset on every image

[nullpainter]

unfortunately there will be a problem for exact FY-2H crop values, as every capture has drift. Sat orbit is unstable which results in xy offset on every image

I guess there's no EXIF metadata to indicate the offset, he says optimistically...

Before I hard-coded crop values, I had code which tried to guess the crop. This approach works okay for IR images. Perhaps I should resurrect it and use it for FY-2H. May be better than nothing?