Category Archives: Image Processing

Astronomical and other image processing using standard software or HALCON.

Waning Gibbous Venus

Posted on by

I finally managed to get a recognizable look at Venus. The gibbous phase is clearly visible as Venus progresses towards its greatest eastern elongation that will occur on November 1st (at that time it will have maximum distance from the Sun and appear “half lit”). As Venus is always hidden under 100% cloud cover, there are no surface details visible even at the best of times.

The image was taken around sunset when the sky was still very bright to catch Venus as high as possible, but even so Venus was already very low in the sky (~10° altitude from 52°N). Due to the short exposure times, the sky appears mostly black anyway. Here is an unprocessed sample frame for comparison:

image_02242_20_32_46_791

As can be seen, imaging through that much atmosphere severly degrades image quality. Turbulence in the air smears the image, and dispersion (due to wavelength-dependency of refraction) turns objects into little rainbows (in fact, for bright objects the atmosphere can be used as a poor man’s spectograph).

However, stacking multiple of these short exposures averages out some of the distortion, while the rainbow effect can be diminished by aligning all color channels individually. I took about 20,000 frames at 0.5 ms exposure over a period of about 4 minutes, 10,000 of which were used for the image presented here.

Saturn is currently located a little to the upper left of Venus (for Nothern Hemisphere observers). It is certainly not a reasonable time to image Saturn, but it was the logical target after Venus:

Six moons and a planet

Posted on by

In the race to take best resolution images of Jupiter and Saturn at highest powers and frame rates, the moons often get neglected. Personally, I do like to see entire systems with their ever-changing view.

Of course, the moons are fainter than the planet so usually the planet will be terribly overexposed when the moons are imaged. At the cost of photometric unscientificness, this can be overcome by merging multiple exposure times in a single image. Here is one of Saturn that shows six of its moons, including faint Hyperion at around 15th magnitude.

And because that is really just a lot of dots without context, here is an annotated version

The process I use is to take a few short exposure frames as for normal planetary images, then one or two longer exposures, then repeat. Each of these frame sets is combined into a meta-frame by some scripting fu that attempts to pull signal from the longer exposures into the dark areas of the shorter exposures. The composite frames can then be stacked as usual. I made a similar one in March for Jupiter

For that one I actually took a gazillion frames over three hours in the hope of making a video out of it once I fully automate the compositing process. Unfortunately I seem to be building up a bit of an image processing backlog. Even now, while I’m trying to reduce the pile, the sky turned clear and the telescope is collecting data outside – argh!