Naked Science Forum

Non Life Sciences => Technology => Topic started by: McKay on 03/04/2014 19:21:31

Title: Analog alternatives for digital cameras replacing mirrors in cars
Post by: McKay on 03/04/2014 19:21:31
I was thinking about cameras replacing mirrors in cars and wanted to think out loud here, if you dont mind:
Cameras use lenses to project the image whatever it is "seeing" to a matrix of light sensitiv pixels that turn it to electric signals and sends them to a monitor to display the image somewhere else, right? Isn't it possible to use the same (or different?)  lenses and, instead of projecting the image to a light sensitive matrix, project it to a mirror, or series of mirrors, to navigate the light/ image to where it can be seen by the driver? That way we would not need a large mirror outside, but could have a smaller "eyeball" outside instead.

Speaking about navigating light (images in particular) around - what other analog ways of navigating light around without destroying the image it carries?
Could optical cables somehow be used? (optical cables do consist of many small mirrors inside .. hmm)
Title: Re: About replacing mirrors with cameras in cars
Post by: CliffordK on 03/04/2014 20:20:41
The original design of the Aptera car had cameras and no mirrors to reduce wind resistance.  Later prototype models had mirrors...  And somehow the car hasn't made it to production yet.  I assume the government nixed the idea.

Certainly the advantage of cameras instead of mirrors is that they can be placed in the optimal location, and potentially also displayed in a more central location in the car so that one keeps one's eyes on the roads rather looking out the windows to try to find the mirrors.  And designed right, they wouldn't get bumped out of focus in the parking lot.  Heck, designed right, and they would never have to be refocused for changing drivers.

Mom's car has backup cameras, but I haven't used it enough to be comfortable with it.  When trying to back into "traffic", one can have sideways facing cameras, as well as back facing cameras (or a fish eye wide angle camera) to provide views that would not be available in the drivers seat.  I'd also put a similar system in the most forward located place (front bumpers) for pulling into traffic around a blind corner.

The displays should be located central to the driver's view.

The problem with an active system (lots of electronics) is that it can fail.  Not that mirrors don't get broken, fall of the windshield, and etc.  If the displays are located in a different location than people are used to looking, then it would require extra learning for the new system.  I'm not sure of the wattage, but one would have to compare system efficiency to wind resistance losses.  Would you place the cameras on small outward facing flexible "antennas" to get a better view to the side?

A passive system also has problems.  Say your collector area is 1 square inch (mighty big), but your display is 6"x6", or 36 square inches.  Then each "pixel" would display 1/36th of the light, and the entire display would be mighty dim.  So, you would have to have a collector perhaps slightly larger than the display.  Feed through several lenses to reduce the size for your fiber optics, then more lenses to return the image to the original size for display.  Would you loose vital clarity and details?
Title: Re: About replacing mirrors with cameras in cars
Post by: RD on 03/04/2014 20:21:22
... what other analog ways of navigating light around without destroying the image it carries?
Could optical cables somehow be used? (optical cables do consist of many small mirrors inside .. hmm)

You'd need an array of cables each the size of a pixel , and the image at the receiver-end would be noticeably dimmer ...

(https://www.thenakedscientists.com/forum/proxy.php?request=http%3A%2F%2Fupload.wikimedia.org%2Fwikipedia%2Fcommons%2Fthumb%2Ff%2Ff4%2FTIR_in_PMMA.jpg%2F320px-TIR_in_PMMA.jpg&hash=e6c19c96eba2dbb6bb51e6f5ff34118d)
http://www.thenakedscientists.com/forum/index.php?topic=46511.msg402324#msg402324
Title: Re: About replacing mirrors with cameras in cars
Post by: JP on 03/04/2014 20:34:29
And if you use imaging optics (lenses/mirrors) instead of fibers, you still have the problem of putting together a decent quality imaging system that is robust to the bumps and vibrations of driving as well as temperature fluctuations, dust/salt, etc.

Wires and electronics tend to be fairly robust to all those factors I mentioned above, so it's generally easier to use them. 
Title: Re: About replacing mirrors with cameras in cars
Post by: syhprum on 03/04/2014 20:57:03
My neighbours Nissan SUV has a camera system which displays a box on the screen for reverse parking I could certainly do with it on my old BMW you can see little in the mirrors when you reverse park and have to rely on the bleepers.
Title: Re: About replacing mirrors with cameras in cars
Post by: evan_au on 03/04/2014 21:21:38
Endoscopes (http://en.wikipedia.org/wiki/Endoscope#Use_of_fiber_optics) for viewing the inside of your body already use fiber optics to carry images. Small size and flexibility is important when viewing your intestines!

50,000 fibers produces a 50,000 pixel image.
The camera chips used in smartphones are small, cheap, rugged, consume little power, and support several megapixels for a better quality image.
They generate several megabits per second of video, which may strain the car's normal data bus; this bus usually carries lower data rates like movement of the steering wheel, accelerator and blinker controls, as well as wheel rotation rates.

Camera chips are also potentially sensitive slightly into infra-red wavelengths, which may help night driving?
Title: Re: About replacing mirrors with cameras in cars
Post by: McKay on 03/04/2014 22:21:02
Quote
You'd need an array of cables each the size of a pixel

But cant a single cable carry the equivalent of multiple pixels? That is, one fiber cable carries only one color/ photon at one angle at a time?

Quote
, and the image at the receiver-end would be noticeably dimmer ...

Hmm, yes. About half as dim?


So, all in all, the main problems with steering light of a image around with fiber optics are: Graininess (lack of pixels) and dimness? 
With carefully aligned rigid system of mirrors: has to be pretty damn rigid and and stable in a vibrating car and, again, dimness?
Title: Re: About replacing mirrors with cameras in cars
Post by: McKay on 03/04/2014 22:29:33
Fiber optics simply work like that? - if I replace my window with a whole lot of short (or long?) fiber optic cables, then I could just see shapes and colors trough the cable wall (albeit dimmer and with some graininess to the image), instead of just white light?
Title: Re: About replacing mirrors with cameras in cars
Post by: RD on 04/04/2014 00:47:06
Endoscopes (http://en.wikipedia.org/wiki/Endoscope#Use_of_fiber_optics) for viewing the inside of your body already use fiber optics to carry images. Small size and flexibility is important when viewing your intestines!

There's PillCam (http://en.wikipedia.org/wiki/Pillcam) to reach the parts that endoscopes cannot reach , (one time use only).

 
Title: Re: About replacing mirrors with cameras in cars
Post by: CliffordK on 04/04/2014 01:50:46
The difference between endoscopes and colonoscopes and a passive fiber optic system being discussed here is that the endoscopes and colonoscopes have active lighting, and may also have a video camera to create a digital display. 

As I mentioned above, if you had a passive system with a 1" x 1" receiver and a 6" x 6" display, then it would be less than 1/36th as bright.  Add another 2:1 factor?  1/72nd?

You'd be much better off using cameras and digital displays that are both already in use in cars.

The Japanese use "fender mirrors" which look like little antennas sprouting out of the front of the vehicle to help with passenger side visibility. 

Apparently Mazda is already introducing an auxiliary camera on the mirror systems (http://www.mlive.com/business/west-michigan/index.ssf/2012/05/gentex_says_its_display_mirror.html)..
Title: Re: About replacing mirrors with cameras in cars
Post by: evan_au on 04/04/2014 13:19:30
When discussing this topic today, a friend today suggested that the function of the rear-vision mirror could be extended, showing a panoramic, all-around view.
In the center, this could show the view out the rear (from a camera); at each side it could show the view from cameras on each side of the car. This view would not be obstructed by pillars, headrests, or passengers.

In terms of usability, people are familiar with glancing up to see out the rear of the car - it is a small step to glance up to see out the rear and both sides.
Title: Re: About replacing mirrors with cameras in cars
Post by: CliffordK on 04/04/2014 20:21:03
About 20 or 30 years ago, they started putting convex mirrors on the passenger side, all with the warning "objects in the mirror may be closer than they appear". 

I agree there would certainly be benefits of having a 180° view from the rearmost portion of the vehicle.  Perhaps add a wing view showing what is directly to your side.  However, the size of the objects would be reduced and detail would be lost.  Thus one would loose much of  the distance perspective, which may or may not be a problem.  Having multiple images in the "mirror" might also be problematic.
Title: Re: About replacing mirrors with cameras in cars
Post by: CliffordK on 04/04/2014 21:55:36
Looking on the web, there is quite a number of rearview mirrors with an integrated display (https://www.google.com/search?q=rearview+mirror+with+%28video+OR+camera+OR+display%29&tbm=isch&sa=X&biw=1094&bih=926).  Most have both forward and rearward video recording capabilities.