23. 5. 2014

Good morning ISS!

I set my alarm clock to 2-3 am to spot the ISS three mornings in a row this week. Today was the day when morning sky looked good and most importantly - I had the mood to get out of the bed.

I prepared Canon 350D with remote controller some days ago. The tripod is from my small astro-telescope.



The email from HQ-spotthestation@mail.nasa.gov said this:

Time: Fri May 23 4:15 AM, Visible: 4 min, Max Height: 66 degrees, Appears: W, Disappears: NE

My location was the road between Bratislava and Kittsee. There I aimed the camera somewhere to the sky. With 18 mm lens (Canon EF-S 18-55mm, f/3.5-5.6) I was not afraid that I miss the flyby of the station. I took some testing shots just to be sure. The sky was quite bright around 4 am therefore I set ISO to 100. The lens focusing was set to manual mode.

I couldn't see the station exactly at 4:15 most probably because of the morning haze. Suddenly it appeared and with full of joy and excitement I pressed the shutter with remote controller. I assume the exposition was around 1.5 min.

After processing in iPhoto and GIMP here is the final image (my first photo of the ISS!):




Texture elimination from scanned photo

Recently I needed to scan quite old photography that was aimed to be published in one local newspaper. I wanted the best quality with dpi around 300. Unfortunately it had this very old-fashioned texture on the top that makes it difficult to scan it properly. The texture will be still there and ruin the result quality.

After some googling about moiré patterns and fast fourier transforms (FFT, it is also used in MRI) I found this very helpful discussion:
http://photo.stackexchange.com/questions/23445/what-is-the-best-way-to-remove-texture-from-a-scanned-textured-photo-paper
- especially the point about FFT.

Here is the technique with FFT:

  1. Download the ImageJ software
  2. Select the part of the image with pattern (e.g. somewhere in the corner). The idea is to see just the pattern.
  3. Apply FFT on the selected image pattern and notice how does it look like in the frequency space.
  4. Apply FFT on the whole image and find the pattern that was present on the selected image after FFT.
  5. Using paintbrush paint the pattern with black color. Do not paint the center.
  6. Apply inverse FFT.

Here is my example:



There is another interesting technique where you scan the image twice with rotation of the second scan about 180°. Then align the images. I didn't use this because the FFT method was much more interesting for me. I would use it probably if I couldn't find the clearly defined pattern in the image (e.g. no clear background).