Author Topic: Episode 705  (Read 4772 times)

0 Members and 1 Guest are viewing this topic.

Offline Ah.hell

  • Poster of Extraordinary Magnitude
  • **********
  • Posts: 13206
Re: Episode 705
« Reply #45 on: January 16, 2019, 02:12:01 PM »
Nevermind. 

Offline werecow

  • Cryptobovinologist
  • Stopped Going Outside
  • *******
  • Posts: 4831
  • mooh
Re: Episode 705
« Reply #46 on: January 16, 2019, 06:55:09 PM »
(click to show/hide)

I had this thought as well.  It could be that getting a few pixels from a lot of directions is easier to process than millions of pixels from a wide-angle camera because you know the exact direction each pixel is coming from.  I also wonder if the compound eyes the want to put on vehicles are for visible light or are lidar or radar.  I don't see how a visible light compound eye would serve any purpose, but a compound radar eye would be awesome.

But lidar and radar are different from vision in that they require one to actively send out and not just passively receive information, so how would that work? And as to processing a few pixels being easier than millions, one can always subsample in any number of ways. And I don't think compound eyes only pick up on photons coming in from a specific angle. So I don't see the advantage there either. Unfortunately the study is behind a paywall. I wish all science journals were open access.
« Last Edit: January 16, 2019, 06:59:32 PM by werecow »
Mooohn!

Offline DamoET

  • Off to a Start
  • *
  • Posts: 99
Re: Episode 705
« Reply #47 on: January 17, 2019, 05:24:29 AM »
Good episode guys, I have been trying to attach some pics of the Magellanic clouds, but no joy.



Damien
« Last Edit: January 17, 2019, 05:27:37 AM by DamoET »
Do not fear God, fear ignorance.  For ignorance will lead you to fearing 'God' (and a whole bunch of other whacky shit!)

Offline daniel1948

  • Isn’t a
  • Reef Tank Owner
  • *********
  • Posts: 8401
  • I'd rather be paddling
Re: Episode 705
« Reply #48 on: January 17, 2019, 09:18:35 AM »
(click to show/hide)

I had this thought as well.  It could be that getting a few pixels from a lot of directions is easier to process than millions of pixels from a wide-angle camera because you know the exact direction each pixel is coming from.  I also wonder if the compound eyes the want to put on vehicles are for visible light or are lidar or radar.  I don't see how a visible light compound eye would serve any purpose, but a compound radar eye would be awesome.

But lidar and radar are different from vision in that they require one to actively send out and not just passively receive information, so how would that work? And as to processing a few pixels being easier than millions, one can always subsample in any number of ways. And I don't think compound eyes only pick up on photons coming in from a specific angle. So I don't see the advantage there either. Unfortunately the study is behind a paywall. I wish all science journals were open access.

I thought that each segment in insect eyes had an extremely narrow field of view, making it easy to detect motion, as objects moved rapidly from one segment to the next. The large number of segments, distributed over a hemisphere, gives an overall field of view of half a sphere in each eye, so with two eyes the insect would be able to see in all directions at once.

This was my impression, anyway.
Daniel
----------------
"Anyone who has ever looked into the glazed eyes of a soldier dying on the battlefield will think long and hard before starting a war."
-- Otto von Bismarck

Offline werecow

  • Cryptobovinologist
  • Stopped Going Outside
  • *******
  • Posts: 4831
  • mooh
Re: Episode 705
« Reply #49 on: January 17, 2019, 11:30:40 AM »
(click to show/hide)

I had this thought as well.  It could be that getting a few pixels from a lot of directions is easier to process than millions of pixels from a wide-angle camera because you know the exact direction each pixel is coming from.  I also wonder if the compound eyes the want to put on vehicles are for visible light or are lidar or radar.  I don't see how a visible light compound eye would serve any purpose, but a compound radar eye would be awesome.

But lidar and radar are different from vision in that they require one to actively send out and not just passively receive information, so how would that work? And as to processing a few pixels being easier than millions, one can always subsample in any number of ways. And I don't think compound eyes only pick up on photons coming in from a specific angle. So I don't see the advantage there either. Unfortunately the study is behind a paywall. I wish all science journals were open access.

I thought that each segment in insect eyes had an extremely narrow field of view, making it easy to detect motion, as objects moved rapidly from one segment to the next. The large number of segments, distributed over a hemisphere, gives an overall field of view of half a sphere in each eye, so with two eyes the insect would be able to see in all directions at once.

This was my impression, anyway.

What's the difference between that and detecting motion across pixels across a camera with a wide FOV?
Mooohn!

Offline brilligtove

  • Too Much Spare Time
  • ********
  • Posts: 7258
  • Ignorance can be cured. Stupidity, you deal with.
Re: Episode 705
« Reply #50 on: January 17, 2019, 12:47:03 PM »
Good question. Any entimologists in the house?
evidence trumps experience | performance over perfection | responsibility – authority = scapegoat | emotions motivate; data doesn't

Offline Billzbub

  • Stopped Going Outside
  • *******
  • Posts: 4250
  • I know you know I know
Re: Episode 705
« Reply #51 on: January 17, 2019, 01:56:25 PM »
(click to show/hide)

I had this thought as well.  It could be that getting a few pixels from a lot of directions is easier to process than millions of pixels from a wide-angle camera because you know the exact direction each pixel is coming from.  I also wonder if the compound eyes the want to put on vehicles are for visible light or are lidar or radar.  I don't see how a visible light compound eye would serve any purpose, but a compound radar eye would be awesome.

But lidar and radar are different from vision in that they require one to actively send out and not just passively receive information, so how would that work? And as to processing a few pixels being easier than millions, one can always subsample in any number of ways. And I don't think compound eyes only pick up on photons coming in from a specific angle. So I don't see the advantage there either. Unfortunately the study is behind a paywall. I wish all science journals were open access.

Yeah, I see your point.
Quote from: Steven Novella
gleefully altering one’s beliefs to accommodate new information should be a badge of honor

Offline arthwollipot

  • Reef Tank Owner
  • *********
  • Posts: 8744
  • Observer of Phenomena
Re: Episode 705
« Reply #52 on: January 17, 2019, 05:30:22 PM »
Fun science fact: The compound eyes of trilobites were made out of silica crystal.
Self-described nerd. Pronouns: He/Him.
Morticia Addams was elegaunt.

Offline The Latinist

  • Cyber Greasemonkey
  • Technical Administrator
  • Too Much Spare Time
  • *****
  • Posts: 7617
Re: Episode 705
« Reply #53 on: January 17, 2019, 06:45:44 PM »
Fun science fact: The compound eyes of trilobites were made out of silica crystal.

This didn’t make much sense to me given what I know about biological systems, so I looked it up.  Turns out that it was calcite, which makes a lot more sense given the biological systems we have for creating various carbonates.
I would like to propose...that...it is undesirable to believe in a proposition when there is no ground whatever for supposing it true. — Bertrand Russell

Offline arthwollipot

  • Reef Tank Owner
  • *********
  • Posts: 8744
  • Observer of Phenomena
Re: Episode 705
« Reply #54 on: January 17, 2019, 06:58:45 PM »
Fun science fact: The compound eyes of trilobites were made out of silica crystal.

This didn’t make much sense to me given what I know about biological systems, so I looked it up.  Turns out that it was calcite, which makes a lot more sense given the biological systems we have for creating various carbonates.

What an embarrassing mistake! Yes, it was calcite. Sorry. It's been a while since I've read the book.
Self-described nerd. Pronouns: He/Him.
Morticia Addams was elegaunt.

Offline daniel1948

  • Isn’t a
  • Reef Tank Owner
  • *********
  • Posts: 8401
  • I'd rather be paddling
Re: Episode 705
« Reply #55 on: January 17, 2019, 06:59:59 PM »
(click to show/hide)

I had this thought as well.  It could be that getting a few pixels from a lot of directions is easier to process than millions of pixels from a wide-angle camera because you know the exact direction each pixel is coming from.  I also wonder if the compound eyes the want to put on vehicles are for visible light or are lidar or radar.  I don't see how a visible light compound eye would serve any purpose, but a compound radar eye would be awesome.

But lidar and radar are different from vision in that they require one to actively send out and not just passively receive information, so how would that work? And as to processing a few pixels being easier than millions, one can always subsample in any number of ways. And I don't think compound eyes only pick up on photons coming in from a specific angle. So I don't see the advantage there either. Unfortunately the study is behind a paywall. I wish all science journals were open access.

I thought that each segment in insect eyes had an extremely narrow field of view, making it easy to detect motion, as objects moved rapidly from one segment to the next. The large number of segments, distributed over a hemisphere, gives an overall field of view of half a sphere in each eye, so with two eyes the insect would be able to see in all directions at once.

This was my impression, anyway.

What's the difference between that and detecting motion across pixels across a camera with a wide FOV?

Just a guess here: with about seven brain cells total, bugs don’t have the processing power. But I think what you’re getting at is why would this system be better for computers? And if you find out and tell me then we’ll both know, because I don’t have a clue.

The whole bug-eye camera thing sounds fishy (buggy?) to me because a camera just exposes pixels on a ccd. What’s the advantage of multiple cameras? Which I think is what you’ve been saying.

I think I understand why it works better for bugs. I cannot imagine it being an I provement for cameras and computers.
Daniel
----------------
"Anyone who has ever looked into the glazed eyes of a soldier dying on the battlefield will think long and hard before starting a war."
-- Otto von Bismarck

Offline CarbShark

  • Poster of Extraordinary Magnitude
  • **********
  • Posts: 11323
Re: Episode 705
« Reply #56 on: January 17, 2019, 07:11:09 PM »
 
Just a guess here: with about seven brain cells total, bugs don’t have the processing power. But I think what you’re getting at is why would this system be better for computers? And if you find out and tell me then we’ll both know, because I don’t have a clue.

The whole bug-eye camera thing sounds fishy (buggy?) to me because a camera just exposes pixels on a ccd. What’s the advantage of multiple cameras? Which I think is what you’ve been saying.

I think I understand why it works better for bugs. I cannot imagine it being an I provement for cameras and computers.

I think the advantage comes from the simple geometry of the lens and receptors. Having a slight difference in the angle that the light hits the receptors may make subtle differences between images, which in itself would make movement easier to detect.  From a single vantage point an object may seem stationary, or it's speed and direction may not be clear, but from multiple angles, even though the parallax factor would be minimal, subtle differences would be more apparent.

That's my guess.
and Donald Trump is President of the United States.

I'm not a doctor, I'm just someone who has done a ton of research into diet and nutrition.

Offline CarbShark

  • Poster of Extraordinary Magnitude
  • **********
  • Posts: 11323
Re: Episode 705
« Reply #57 on: January 17, 2019, 07:12:27 PM »
Fun science fact: The compound eyes of trilobites were made out of silica crystal.

Don't they become something like silica crystal after a few million years in the Lyme Regis cliffs?
and Donald Trump is President of the United States.

I'm not a doctor, I'm just someone who has done a ton of research into diet and nutrition.

Offline arthwollipot

  • Reef Tank Owner
  • *********
  • Posts: 8744
  • Observer of Phenomena
Re: Episode 705
« Reply #58 on: January 17, 2019, 07:32:07 PM »
Fun science fact: The compound eyes of trilobites were made out of silica crystal.

Don't they become something like silica crystal after a few million years in the Lyme Regis cliffs?

No, that's the amazing thing. Given the mistake I made and that they're calcite crystal rather than silica. But the living animals themselves had these calcite eyes. There's a LOT more detail in that wonderful book I linked to in my last post if you're interested.
Self-described nerd. Pronouns: He/Him.
Morticia Addams was elegaunt.

Online bachfiend

  • Not Any Kind of Moderator
  • Well Established
  • *****
  • Posts: 1680
Re: Episode 705
« Reply #59 on: January 17, 2019, 08:33:54 PM »
Just a guess here: with about seven brain cells total, bugs don’t have the processing power. But I think what you’re getting at is why would this system be better for computers? And if you find out and tell me then we’ll both know, because I don’t have a clue.

The whole bug-eye camera thing sounds fishy (buggy?) to me because a camera just exposes pixels on a ccd. What’s the advantage of multiple cameras? Which I think is what you’ve been saying.

I think I understand why it works better for bugs. I cannot imagine it being an I provement for cameras and computers.

I think the advantage comes from the simple geometry of the lens and receptors. Having a slight difference in the angle that the light hits the receptors may make subtle differences between images, which in itself would make movement easier to detect.  From a single vantage point an object may seem stationary, or it's speed and direction may not be clear, but from multiple angles, even though the parallax factor would be minimal, subtle differences would be more apparent.

That's my guess.

I think that in the discussion of compound eyes on the show the idea that each lens is providing a complete image of the external world was dismissed as not being true.  Each lens, with its photoreceptor, provides a pixel in the eventual image, in the same way that fibreoptic endoscopes work - the fibreoptic endoscope consists of bundles of very fine glass filaments each providing a very small part of the image.

An object moving through the visual field of a compound eye would stimulate new photoreceptors as it stops stimulating others, so the motion would be detected.  At one moment, photoreceptors ABC might be firing, then BCD, followed by CDE.
Gebt ihr ihr ihr Buch zurück?

 

personate-rain