An analysis of the Betty Hill "star map"

originally posted by: Keiyentai
a reply to: james1947

Celestia is good though if you know Python and want good 3D models/renders use Blender. It's open source and besides the horrid UI which hasn't been changed since day one practically -_- it out does Poser 11 Pro in every aspect. Poser is mainly for animating human models and basic 3D primitive models and the render engine in 11 Pro is dated. Been using Poser since Bryce/Corel owned it before Sun Smith bought it from DAZ. Also interesting thing is a few years ago a person made a map based off the Betty Hill map in Poser and even in Poser 14 it was...bad even with the "new render engine" in 14 Pro. Would be interesting to see the 3D map imported to Blender. Good work though.

You guys are confusing the bleep outta me, talkin about all kinds of, R2-D2s and C-3POs over here...

originally posted by: Arbitrageur

originally posted by: james1947

Betty's map cleaned up

My map...

Those two images don't look that much alike to me. I think you are biased to want to believe in the Betty Hill star map.

The software methods were developed by AForgeNet, they reported 99.1%. IF there is any bias, it is not mine...
Frankly, I do see quite a lot of similarity between these two images, I even did a "manual blob analysis" it came out to a more modest 82.5%, but, that is a rather crude method.

originally posted by: james1947

The software methods were developed by AForgeNet, they reported 99.1%. IF there is any bias, it is not mine...
Frankly, I do see quite a lot of similarity between these two images, I even did a "manual blob analysis" it came out to a more modest 82.5%, but, that is a rather crude method.

Thank you for the thread, James - it didn't seem logical to go into too much depth in the other "acronym" thread, so I'm glad you took up the suggestion of having your own thread to 'captain'.


I must admit I needed a few strong coffees to tackle your complex thesis, but I'm getting there slowly!

By the way, am I imagining it or did Betty say she had natural recall regarding the map, without the aid of regression - albeit not in such detail, of course? (I know Barney remembered virtually nothing of the craft's interior without regression.)

a reply to: james1947

A few points. Quotes are from Betty Hill in the Interrupted Journey or the Hill Report.

- Betty Hill never originally described the map in 3 dimensions. It was added to the story several years later in 1969 when Majorie Fish could only make a match using a 3-D model she made. Then it was described as "looking through a window" via Fish's description of what she said. If Betty would have seen something as amazing as a 3-D map in 1961, she would have mentioned it in her story. She did not.

He went over to the wall and pulled down a map. This chart..."

- Betty drew curved lines between points which is a classic way to represent distance on a 2-D human map.

they were not straight, but curved.
.
tiny dots connected by curved lines

- The leader told Betty that the solid lines were trade routes and the dashed lines were exploration routes. These are both human terms used to describe maps such as the silk road etc.

- Betty drew curved lines on two of the closest points. That's more than likely to represent the curvature of a planet, not two gas stars. That's also consistent with her quote saying her map had planets.

- Why were the aliens only visiting stars according to your map? Hot star hopping to trade goods and explore?

- Why would our sun be marked with solid lined trade route when the visit is described as a meeting and exploration of humans? The lines should be dashed/dotted as exploration routes of our solar system.

- Why would the alien leader ask Betty where Earth is on the map if the map only showed stars? "Where is your sun?" is a more logical question when Earth is not even shown.

he asked me where earth was on this map

- The entire abduction portion, including the map, was a part of several dreams that Betty had. Dreams that she wrote in her journal over several days in September 1961, 2 years before her hypnosis and 5 years the book was written. Page 297 in The Interrupted Journey.

- Betty described both stars and planets on her map:

it was a map of the heavens, with numerous size stars and planets
.
Some went from one planet to another, to another

Your map only shows stars. Where are the planets that Betty described?

These are some questions that need to be addressed even before the map can be taken seriously. You also say her map is only a representation of what she saw. Doing it this way allows you to move the points and lines on your map and look at different perspectives in 3-D space to get a match. Once you start moving things, it can no longer can be said to be a match. At best, you can say it resembles Betty's map. Everyone can see this james1947 and it might be one reason you can't find support of your map here and elsewhere after several years of trying. From a visual stand point there's no match. If there was a match, you may have a valid argument.
A for effort -- F for execution.

By the way, by association with this "factual" map, these become facts also:
- Some aliens had dark hair with large Jimmy Durante noses.
- The leader alien spoken English with a foreign accent.
- Some aliens wore light blue trousers and zipper style of sport jackets.
- Some had slip on boots.
- Military caps were worn by many.
- Barney described the spacecraft in the stereotypical 50's and 60's movie/TV show style. There was a long row of windows, a control panel with levers on it can be seen inside, a ramp lowered from the bottom to pick up Betty and Barney, aliens can be seen walking about inside the craft, etc. He also goes on to describe an angry alien watching him from inside wearing a black shiny leather outfit, a black scarf, and a cap with a bill on it. A friendlier alien who looks over his right shoulder from inside the craft and smiles at him. He has a round face and reminds him of an Irishman.
All sounds very WWI pilot and Hitler-like and human.

This above is not to mention the kooky behavior and UFO nut Betty revealed herself to be.

I think Betty sat at her kitchen table and drew a map she believed would convince people using hints of human artistic style to it. You see it in her drawing. Who knows what photo or drawing she may have used as a visual reference and inspiration. I showed a possibility in another thread, so did someone else in this one. Using curved lines to show distance, dashes and solid lies to assign different meanings, and mentioning of trade and exploration routes are human additions to maps too common to be coincidental.

originally posted by: Ectoplasm8
a reply to: james1947

Wow...where to start.

A few points. Quotes are from Betty Hill in the Interrupted Journey or the Hill Report.

- Betty Hill never originally described the map in 3 dimensions. It was added to the story several years later in 1969 when Majorie Fish could only make a match using a 3-D model she made. Then it was described as "looking through a window" via Fish's description of what she said. If Betty would have seen something as amazing as a 3-D map in 1961, she would have mentioned it in her story. She did not.

He went over to the wall and pulled down a map. This chart..."

Actually, I believe I mentioned that the displays probably were not holographic, it is possible to represent 3D space on a 2D surface, it's actually done all the time...World of Warcraft comes to mind. But, you're right she probably wouldn't notice any difference, or much anyway, from television displays of similar content.

- The leader told Betty that the solid lines were trade routes and the dashed lines were exploration routes. These are both human terms used to describe maps such as the silk road etc.

I would think that both of these terms/ideas would be Universal in scope, and translatable into virtually every language so-ever.

- Betty described both stars and planets on her map:

it was a map of the heavens, with numerous size stars and planets
.
Some went from one planet to another, to another

Your map only shows stars. Where are the planets that Betty described?

Actually, my map does show both stars and planets, though perhaps not as well as I can make it...but then it is version 1.0, lets see what I can do for 2.0. Though the "planets" part might not be as pretty.

Go to the website. If you mouse over a star it will show you its name, if you click on a star it will show you some astrometrics, AND, any planets that have been discovered there.

You can also navigate to the PDF and see the planets, though, those are the same planetary images that are in the GUI.
though, that PDF is a draft and subject to change.

These are some questions that need to be addressed even before the map can be taken seriously.

Many of your questions are more than a little outside the scope of this analysis.
For instance; I'm not addressing the mindset of any individual, nor do I address the cause of any behavior. This analysis only addresses the map, but you already knew that.

You also say her map is only a representation of what she saw. Doing it this way allows you to move the points and lines on your map and look at different perspectives in 3-D space to get a match. Once you start moving things, it can no longer can be said to be a match. At best, you can say it resembles Betty's map. Everyone can see this james1947 and it might be one reason you can't find support of your map here and elsewhere after several years of trying. From a visual stand point there's no match. If there was a match, you may have a valid argument.

Yes, it is only a representation of what she saw, it can't anything other. However, you might want to re-think that bit about who is "moving points" around. You see, Betty saw a map of real stars, and then drew it after some time; so her "drawing" can't be all that accurate. It was Betty that was "moving points".

And, yes, in a sort of technical way, we can call it a match. Did you notice the "percentage" part? That is the quality of the match. And, that percentage actually makes the probability of coincidence greater, even though it is meaningless.

A for effort -- F for execution.

Actually man, I didn't put this out here for a "grade"

By the way, by association with this "factual" map, these become facts also:
- Some aliens had dark hair with large Jimmy Durante noses.
- The leader alien spoken English with a foreign accent.
- Some aliens wore light blue trousers and zipper style of sport jackets.
- Some had slip on boots.
- Military caps were worn by many.

Well, actually; no, there is no supporting data for any of that.

I think Betty sat at her kitchen table and drew a map she believed would convince people using hints of human artistic style to it. You see it in her drawing.

I would think that rather improbable, actually "vanishingly small"! There is still the issue of four (4) stars that are predicted by the map, and weren't cataloged until 1992.

a reply to: james1947

How do you feel about details Bob Lazar gave, before they were publicly considered factual and backed by data?

Some details he gave out, ended up very accurate, years later.

Also, if you like to play the odds game on this kind of stuff, you might find the FRB data of interest as well.

If you're going to dabble in this kind of stuff, there are other data sets you can look at, for probability anomalies.

What else, can you use your talents on? I'm interested. You're very meticulous, there are probably other things posted on this site, your type of work would benefit.

a reply to: Archivalist

Some details he gave out, ended up very accurate, years later.

Nuh uh.

I think your efforts are very commendable. This is the sort of thing we need more of when trying to get to grips with these kind of phenomena.

A very good job.

Have you been able to identify any weaknesses in your method or approach? Any potential ways to improve upon it?

What if you generate a set of say 10 random dots and see what your best match is, and an average?

originally posted by: james1947
Yes, it is only a representation of what she saw, it can't anything other. However, you might want to re-think that bit about who is "moving points" around. You see, Betty saw a map of real stars, and then drew it after some time; so her "drawing" can't be all that accurate. It was Betty that was "moving points".

If we cannot know the accuracy of Betty's map this is all useless, you may have a high accuracy for something completely wrong.

originally posted by: james1947

originally posted by: Arbitrageur

originally posted by: james1947

Betty's map cleaned up

My map...

Those two images don't look that much alike to me. I think you are biased to want to believe in the Betty Hill star map.

The software methods were developed by AForgeNet, they reported 99.1%. IF there is any bias, it is not mine...
Frankly, I do see quite a lot of similarity between these two images, I even did a "manual blob analysis" it came out to a more modest 82.5%, but, that is a rather crude method.

Do you want to try an experiment? I made a fake map with 12 stars, a group of 3 in each corner. What does AForgeNet say about how well my fake map matches yours? What number do you get?

Here is my fake map if you want to run the experiment:



I have software that does something similar to what AForgeNet does, and it shows a 99.5% match of your map to my fake map:



I circled the 99.5% match result.

From the documentation on the tool you used:

www.aforgenet.com...

The class implements exhaustive template matching algorithm, which performs complete scan of source image, comparing each pixel with corresponding pixel of template.

That is more or less what my software tool did and it found 99.5% of the pixels are identical between my fake map and your real map, which isn't surprising since most of them are white, so of course a high percentage will match no matter where the stars are drawn.

a reply to: Arbitrageur

Okay, I tried you experiment, after all that is what science is about.

No match

"From the documentation on the tool you used: www.aforgenet.com... The class implements exhaustive template matching algorithm, which performs complete scan of source image, comparing each pixel with corresponding pixel of template. "

That is more or less what my software tool did and it found 99.5% of the pixels are identical between my fake map and your real map, which isn't surprising since most of them are white, so of course a high percentage will match no matter where the stars are drawn.

I have my program defaulted to 99%, so any match that is less won't be found, and 6+ minutes of working; no match was found. An image your size typically takes 6 - 8 minutes.

Then I zoomed in on your dialogbox image...Dude you are looking at individual pixels, as contrasted with a "Template".

We are not doing a pixel comparison here, that won't work. Rather we are comparing a collection of "dots" to another collection of "dots".

You made another serious mistake; what is the name and producer of the software you used?

I'll be completely honest with you; I wrote 99% of the software required for this project. That would be all of it, except for any library classes I may have needed, like the AForge classes.

originally posted by: ArMaP

originally posted by: james1947
Yes, it is only a representation of what she saw, it can't anything other. However, you might want to re-think that bit about who is "moving points" around. You see, Betty saw a map of real stars, and then drew it after some time; so her "drawing" can't be all that accurate. It was Betty that was "moving points".

If we cannot know the accuracy of Betty's map this is all useless, you may have a high accuracy for something completely wrong.

Betty's map, as a stand alone object, doesn't have an accuracy. It can't since it is a hand drawn image that was taken completely from memory. The very best Betty's map can say is; "it was something like this."

Thus is may be incumbent upon us to determine that "accuracy". My software says: 99.1%

This isn't the best starting place, but, sometimes we have to work with what we have.

originally posted by: beetee
I think your efforts are very commendable. This is the sort of thing we need more of when trying to get to grips with these kind of phenomena.

A very good job.

Thank you.

Have you been able to identify any weaknesses in your method or approach? Any potential ways to improve upon it?

What if you generate a set of say 10 random dots and see what your best match is, and an average?

Weaknesses; Not so much with the method, that's kind of technological, and something I can easily grasp. The approach; while I don't really see any weakness there either, it is an area where there may be other, better, ways of analysis. If I should stumble upon them I will probably apply them.

Actually, I've dome something like that, though, I didn't generate any "source images" (that would be your 10 or so dots). I have submitted Arbitrageur's image (no match), and some others, the best so far is the "Hill Wilson map" which matches at 98.6%.

I've been looking for other attempts at interpreting Betty's map, but haven't found much...

a reply to: james1947

I have my program defaulted to 99%, so any match that is less won't be found, and 6+ minutes of working; no match was found.

So the match could be 98.9% and you wouldn't find it? That doesn't seem like a fair test.

We are not doing a pixel comparison here, that won't work.

That's what this reads like to me:

"comparing each pixel with corresponding pixel of template"

There's some kind of huge language barrier here if you say that "comparing each pixel with corresponding pixel of template" is "not doing a pixel comparison".

originally posted by: Arbitrageur
a reply to: james1947

I have my program defaulted to 99%, so any match that is less won't be found, and 6+ minutes of working; no match was found.

So the match could be 89.9% and you wouldn't find it? That doesn't seem like a fair test.

We are not doing a pixel comparison here, that won't work.

That's what this reads like to me:

"comparing each pixel with corresponding pixel of template"

There's some kind of huge language barrier here if you say that "comparing each pixel with corresponding pixel of template" is "not doing a pixel comparison".

Uh-huh...you did read the bit about a Template, right?
I seriously do not want to get into a shouting match about computer vision and A I. It won't be productive.

But, you can not do a simple pixel comparison of two unequally sized images and expect any serious results when One should be doing a template match ...

The two are vastly different...

originally posted by: james1947

originally posted by: Arbitrageur
a reply to: james1947

I have my program defaulted to 99%, so any match that is less won't be found, and 6+ minutes of working; no match was found.

So the match could be 89.9% and you wouldn't find it? That doesn't seem like a fair test.

We are not doing a pixel comparison here, that won't work.

That's what this reads like to me:

"comparing each pixel with corresponding pixel of template"

There's some kind of huge language barrier here if you say that "comparing each pixel with corresponding pixel of template" is "not doing a pixel comparison".

Uh-huh...you did read the bit about a Template, right?
I seriously do not want to get into a shouting match about computer vision and A I. It won't be productive.

But, you can not do a simple pixel comparison of two unequally sized images and expect any serious results when One should be doing a template match ...

The two are vastly different...

I made the test image the same size so the images are not unequally sized in the case of my test image, so it can directly compare pixels.

Yes if the images are different sizes you would need some kind of scaling but still in that case if most pixels are white you'll get a high match because of all the white pixels, no matter where the black pixels are.

originally posted by: Archivalist
a reply to: james1947

How do you feel about details Bob Lazar gave, before they were publicly considered factual and backed by data?

Didn't follow ole Bobby that closely. Though from what I did "hear" wasn't very good, and the most significant part was the whole element 115 thing, which was proven inaccurate when element 115 was produced. That, for me, was rather significant.

Also, if you like to play the odds game on this kind of stuff, you might find the FRB data of interest as well.

FRB's given the rate at which electromagnetic fields degrade over distance; ALL FRB's are natural phenomenon, as contrasted with an artificial one.

Radio (electromagnetic fields) transmitted from Earth are not detectable / distinguishable from background at around 5 light years, and that would be with Earth's very best technology today, and that's technology. An FRB contains more energy than a super nova.

originally posted by: Arbitrageur

originally posted by: james1947

originally posted by: Arbitrageur
a reply to: james1947

I have my program defaulted to 99%, so any match that is less won't be found, and 6+ minutes of working; no match was found.

So the match could be 89.9% and you wouldn't find it? That doesn't seem like a fair test.

We are not doing a pixel comparison here, that won't work.

That's what this reads like to me:

"comparing each pixel with corresponding pixel of template"

There's some kind of huge language barrier here if you say that "comparing each pixel with corresponding pixel of template" is "not doing a pixel comparison".

Uh-huh...you did read the bit about a Template, right?
I seriously do not want to get into a shouting match about computer vision and A I. It won't be productive.

But, you can not do a simple pixel comparison of two unequally sized images and expect any serious results when One should be doing a template match ...

The two are vastly different...

I made the test image the same size so the images are not unequally sized in the case of my test image, so it can directly compare pixels.

Yes if the images are different sizes you would need some kind of scaling but still in that case if most pixels are white you'll get a high match because of all the white pixels, no matter where the black pixels are.

I'm sorry man...but, the background is not a part of the template. Thus background is not considered.

By the way, a largish sourceimage as compared to the template image seem to work better.

Also, IF what you are attempting were even remotely true/correct; then Facial recognition wouldn't work, neither would Optical character recognition, and, well, most of the whole Computer Vision area of technology. Computer Vision is kind of dependent on being able to recognize objects and process the acquired data.

originally posted by: james1947
Betty's map, as a stand alone object, doesn't have an accuracy. It can't since it is a hand drawn image that was taken completely from memory. The very best Betty's map can say is; "it was something like this."

Thus is may be incumbent upon us to determine that "accuracy". My software says: 99.1%

Could you explain it better, please? What "accuracy" are you talking about? Accuracy of Betty's map compared with what?

This isn't the best starting place, but, sometimes we have to work with what we have.

As you know, garbage in, garbage out. Using bad data is the best way of getting bad results.

a reply to: james1947

We are all at a disadvantage in this discussion, as we (at least I, I suppose that applies to most people reading this thread) do not know exactly what you mean by "template" and how your system really works, so could you please explain it better?

For example, you say that you created "a C# application to apply SQL search criteria (everything < 33 parsec)". Less than 33 parsec from what?

Also, you used Poser to create the 3D model. I don't know how poser really works, but 3D programs, usually, have a camera function that replicates common camera configurations for close filming, not space distances, so I don't know if the camera you chose introduced any distortion or not, but I think it's a possibility.

You also say that you removed the second “zeta Reticuli” from Betty's map. Why?
And that the computer generated view was "filtered to show only stars of interest". What stars did you remove and criteria did you use to say they were not of interest?

Finally, how did you test your system? Did you use another star map, showing known stars, to compare with a computer generated model?