New Moon

New Moon: 6 / June / 2019

No filter.
Gear: Redmi S2

On the very first image of black hole: Part II

Related to the first image of a black hole, I explained my opinion in my first post: On the very first image of black hole: Part I
Here I'll try to expand it with an example.
Assume you are trying to take a photo of a billboard far away. You have the information that there is only one word on the billboard. You are limited to digital cameras. If you use only one digital camera you will be able to only capture a very small area of the billboard. The red box in Fig. 1.

Fig. 1: A billboard far away, divided to 56 sections

It is obvious that if one uses 56 cameras and each pointed towards a specific section in Fig. 1, by combining all the captured photos, the real image of the billboard can be revealed. 

The main challenge in 'Event Horizon' project was that they have tried to capture the image of the billboard in Fig. 1 with only 8 cameras. This the area that has been captured: 

Fig. 2: Captured areas with 8 digital cameras

Consider the image of the black hole as a word. What has been done is that all meaningful words has been added to a database. With the characters obtained in Fig. 3, random words has been generated:

Fig. 3: Obtained characters with 8 cameras

Random words:

- Banejele

- Brnersa

- Bqnvsdf

- Bvnoidas

- Banejele

...

...

- Bankrupt

Then these words have been tested with the database, and the word with a meaning has been selected as the image of the billboard (black hole). Which in this case the selected word, for instance, can be:

Bankrupt

What I'm tried to say is that the obtained final word, as the real image of the billboard, is completely related to our database. Here, as my database was an English dictionary, the word that obtained was Bankrupt, however the real image of the billboard is:

Fig. 4: The real image of the billboard

In conclusion, if one day you heard that the image of black hole is different from what you have seen today, dont be supersized. 

On the very first image of black hole: Part I

I was utterly amazed by the first-ever image of a real black hole which was released recently. The captured supermassive black hole is located in the center of the galaxy M87, 55 million light-years from Earth. Due to the size and power limitations, currently, no single-dish telescope can image the black hole. Therefore, rather than one telescope, as described by Katie [2], the first picture of the black hole (Fig. 1) was captured in the Event Horizon Telescope (EHT) project with a network of eight linked together telescopes creating a computational telescope the size of the Earth.

Fig. 1: The first picture of a black hole [1]

The EHT project can be summarized as:

1. The area of interest (where the black hole is located) is expressed as n-by-m mesh creating n*m sections
2. Each telescope has recorded astronomical signal from one of the sections. 
3. Considering the real image of the black hole as n*m-piece puzzle, now 8 (the number of telescopes in the project) pieces are known. 
4. A random image containing the available 8 pieces is generated.
5. An artificial intelligence algorithm, compares the generated image with real images. If the randomly generated image satisfies real image properties it is considered as the image of the black hole, if not, the process is repeated from step 4. 

A variety of images have been added to the database to insure that the algorithm can distinguish a fake image from a real image. Moreover, in EHT project, imaging algorithms with additional assumptions and constraints have been used to produce images that are physically plausible [1].

As described in the related published article [1], the model tries to search for an image that is not only consistent with the observed data but also favors specified image properties.

 Assuming a puzzle with maximum 10-20 pieces, simply, with any 8 pieces the real image can be generated with a high degree of accuracy. The challenge here is that the black hole is a puzzle with millions of pieces and out of that million pieces we now only have 8 pieces (Fig. 2)!

Fig. 2: 8 peaces of Fig. 1

In my opinion, it has not be considered that all the images in the database that has been used to distinguish fake images from real ones, have a common property: They all have been taken by humans.

This common property can significantly affect the image that has now been considered as the first ever picture of a black hole. Putting it differently, assume we use the drawings of 3-4 years old kids as our concept of a real images. The image of the black hole will be different, and will represent a black hole in terms of the drawings of a 3-4 years old kids.

What I’m trying to say is that with the laws of physics that we know and with the concepts that we have on the Earth the image of the black hole will look like Fig. 1, but if we accept that the laws of physics can be different elsewhere (i.e. in the black hole), the image will be different. Or even, with the same concepts and process, taking a picture of the same black hole from another planet, by another intelligent creatures will for sure result differently, as the generated image of the black hole is completely dependent of what we consider as a real or fake image!

To be continue...

References
1- Alberdi, A., Gómez Fernández, J. L., & Event Horizon Telescope Collaboration. (2019). First M87 Event Horizon Telescope Results. I. The Shadow of the Supermassive Black Hole.

2- https://www.ted.com/talks/katie_bouman_what_does_a_black_hole_look_like?language=en