SHG (spectroheliography) software v3: batch pixel shift processing feature added

The place for all your solar software; capture, stacking, post processing, ancillaries etc. It's all up for discussion here!
Post Reply
thesmiths
Almost There...
Almost There...
Posts: 1053
Joined: Mon Jun 04, 2012 8:13 pm
Location: London, England
Has thanked: 42 times
Been thanked: 1485 times

SHG (spectroheliography) software v3: batch pixel shift processing feature added

Post by thesmiths »

Based on user feedback, we have implemented high-speed batch processing of pixel shift into our program. Since each shift pixel corresponds to a wavelength shift, this allows generating a sequence of wavelength shifted images. The computation and generation of, for example, 20 different wavelengths is now done automatically and in about the same time as for a single image.

As before, the software is available from our GitHub: https://github.com/thelondonsmiths/Solex_ser_recon_EN

The GUI looks the same as before:

SHG software GUI
SHG software GUI
pixel GUI.JPG (35.23 KiB) Viewed 1410 times

The README file describes the additional features:
For no shift, leave the "Pixel offset" box at the default of '0'
Specify the output of a particular shift by entering a single number or particular values with commas: 'a,b,c,d,e'
For a range x to y with an interval of w, use colons: 'x:y:w'
If 'w' not specified, the default is 1 so 'x:y' will produce the range x, x+1, x+2, ... y-2, y-1, y
x, y, a, b, c can be positive or negative integers; the number w can only be a positive integer
Batch pixel shift processing of a batch of files is allowed
In general, I would recommend leaving the graphics turned off. Also, I would recommend (at least at first) to use "Save CLAHE.png only" because the number of files written quickly becomes large and confusing. Typically, I have found w = 2 for hydrogen alpha and w = 4 for calcium H works nicely (this assumes single binning on a 178MM; for double binning, a better setting would presumably be w = 1 for hydrogen and w = 2 for calcium).

Typically, the images [-x, 0] look about the same as the ones [0, +x], so I would recommend (at least at first) to do pixel shift only in one direction to avoid a confusing accumulation of output files.

Below is an example of the output files done with the Solex website file 16_33_15 for hydrogen alpha done with the GUI settings show above:

H-alpha pixel shift
H-alpha pixel shift
pixel shift 0 14 2.JPG (83.67 KiB) Viewed 1410 times

Below is an example of a calcium H file I took recently:

Ca-H pixel shift
Ca-H pixel shift
pixel shift 0 28 4.JPG (57.02 KiB) Viewed 1410 times

Here are images from an old hydrogen alpha file I took assembled into a GIF. One advantage of this batch processing method is the shifts are all processed essentially identically so that making an animation is simpler (since the geometry correction is identical). These images were produced using the entry 0:8:2

H-alpha pixel shift GIF
H-alpha pixel shift GIF
170116_shift=0 2 4 6 8.gif (1.07 MiB) Viewed 1410 times

Finally here is a GIF animation of the protuberance images done at -2 and +2 pixel shift. Presumably, the difference is an indication of Doppler shift.

H-alpha prom doppler shift
H-alpha prom doppler shift
170116_shift=-2 2.gif (86.14 KiB) Viewed 1410 times


thesmiths
Almost There...
Almost There...
Posts: 1053
Joined: Mon Jun 04, 2012 8:13 pm
Location: London, England
Has thanked: 42 times
Been thanked: 1485 times

Re: SHG (spectroheliography) software v3: batch pixel shift processing feature added

Post by thesmiths »

By the way, another thing that the batch processing of pixel shift allowed was generation of two "supplemental" images at pixel shift 0 and 10, no matter what the "Pixel Offset" input box is. Pixel shift = 0 is always used to generate the "live view" if graphics is turned on since this typically has the most interesting structure. Conversely, pixel shift = 10 produces what is essentially a "solar continuum" image, no matter what wavelength you are looking at. This has the advantage of being fairly featureless so is much easier to process for edge detection and therefore for geometry correction. Below is the geometry correction window for a Calcium image (sent to me by Valerie) where we have pixel shifted in order to find the edges better. In the traditional method of using the centre of the Calcium spectral line, the contrast between the solar disk and the sky is quite low (and the surface also has a lot of "confusing" surface features).

pixel shifted geometry correction
pixel shifted geometry correction
10_57_39 new.JPG (58.15 KiB) Viewed 1384 times
By essentially converting every spectral line into a "white light" image, the edge detection is now remarkably improved and robust. Because of the batch pixel processing feature, we are able to generate the two additional pixel shift images with almost no time penalty.


mconsidine
Oh, I get it now!
Oh, I get it now!
Posts: 22
Joined: Sun Aug 21, 2016 10:57 pm
Been thanked: 13 times

Re: SHG (spectroheliography) software v3: batch pixel shift processing feature added

Post by mconsidine »

Fwiw, i have updated my version of thelondonsmiths code to include Avi support:
Http://GitHub.com/mattconsidine/solex_ser_recon_avi

There is also a highly experimental version with support for 8bit SER and 8bif AVI files, along with the ability to specify a bandwidth average (similar to Wah's SLiM). There is also an automatic creation of an animated gif if pixel shift covers a range.
Http://GitHub.com/mconsidine/Digital_SHG

Matt


Post Reply