OUR NETWORK:TouchSmart Community TechLore Sling Community MyOpenRouter MediaSmart home See all... About UsAdvertiseContact Us

 
Learn about scoring Forum's Raw Score: 1629180.0
February 7, 2008 08:01 PM

Categories: General Robotics Roboquad

Rating (2 votes)
  • 1
  • 2
  • 3
  • 4
  • 5
Rate This!

Member Avatar

Robo BG

Member
Joined: 01/03/2008

Hi Robocommunity!

  I play with my roboquad a lot.  I was at a friends house and we were making roboquad navigate through a maze.  My friend has these black colored sofas.  During his Escape walk and when he just walked forward he never recognized the sofa was there until his eyes were half a centimeter from touching it.  It happened at my house too with the black leg to my fooseball table.

So I am wondering if maybe the color black somehow absorbs the infared waves.  Like how it absorbs sunlight.  Maybe it absorbs the IR when it is transmitted.  What do you think?

Keep the robotic dreams,
-BG
http://www.robocommunity.com/article/12811/Cool-Programs-To-Use-For-Roboquad-...

Discussion:    Add a Comment | Comments 1-25 of 41 | Latest Comment | 1 2 Next »

February 7, 2008 8:07 PM

Indeed, black messes with IR object detection. The problem is that black is technically a complete lack of color. Other colors are visible as their color because they reflect that color light; for example, red objects reflect red light, white objects reflect all light, black objects reflect no light.

This is the main reason you'll see a lot of homemade robots with IR _and_ sonar sensors. The sonar can "see" black things, the IR can see soft things that don't bounce sound waves well.

Fun stuff, eh? 

Watch out, don't step in the anthropomorphization.

February 7, 2008 8:14 PM

I knew I was right!

I thought black was the presence of all colors and white was the absence? 

Keep the robotic dreams,
-BG
http://www.robocommunity.com/article/12811/Cool-Programs-To-Use-For-Roboquad-...

February 7, 2008 8:23 PM

Nope, black is the lack of colour.

White is the combination of colours. Check this out:

http://en.wikipedia.org/wiki/RGB_color_model

See how the combination of red,green, and blue creates white. Neat eh.

February 8, 2008 5:16 AM

Black is the lack of color when you are using Additive color model used when you are
producing light. Black is all colors when you are using Subtractive color model. 

http://en.wikipedia.org/wiki/Subtractive_color

The difference is a monitor uses Additive, a painting uses Subtractive. Does the item produce light or reflect?

Kanaswind 

February 8, 2008 5:53 AM

If I am recalling my physics correctly, what we see is reflected light. Objects that look a particular colour are actually absorbing the other colours, and only reflecting the ones we see. Black objects are absorbing most wavelengths and reflecting very little. White objects absorb almost no visible wavelengths.

Which is why a white car is cooler than a black car. 

RoboGuide - Your guide to hacking all things WowWee

February 8, 2008 8:55 AM

Nocturnal said: Which is why a white car is cooler than a black car. 

Only in a thermodynamics sort of way.... Wink

As mentioned there is additive and subtractive color changes done by substances. If you take two colors of paint, and mix them to create a third, it is subtractive tinting. Each paint tint absorbs different wavelengths (colors).  If you take the same two colors, but this time as colored light, and shine them together on a white surface, you won't get the same tint that you got with the mixed paint.  Shining two lights together is adding wavelengths to the mix, rather than the way paint mixes absorb additional wavelengths. 

I learned about this in my physics classes, but there's lots of web references available that discuss it. For example,

http://home.att.net/~RTRUSCIO/COLORSYS.htm 

What's this mean?

It means a black surface is a surface that absorbs all visible light wavelengths. As the fabric fades, and turns "grey", it is losing its absorbing properties and reflecting more of each wavelength. 

Conversely, a white surface is one that reflects all visible light wavelengths equally well. If white fabric is washed with a red towel, and turns pink, it means that that the fabric has lost its ability to reflect all wavelengths equally, and that the red wavelengths are reflected more than the blue or green ones.

Infrared is not a visible wavelength. It's a longer wavelength than what the human retina can normally detect. Just because we can't see IR, doesn't mean that it doesn't get reflected and absorbed like other light. It works very much like visible light when it comes to absorbtion and reflection.

Dark black isn't necessarily the best  IR absorber. It has to do with the nature of the fabric, too.

I have a charcoal grey chamois shirt that absorbs a lot of IR. It absorbs more IR than a much blacker tightly woven linen shirt.

You might ask how I know that, but you also might be sorry you did. The practical experiments  involve certain public plumbing fixtures, and the IR detection they use to "arm" themselves for action upon your departure. Nuff said.

ScottE -- Member (always) & Moderator (when needed)

February 8, 2008 2:55 PM

I ran an experiment on two pieces of Construction paper.  He stopped when he saw them at the same mark on 4/5 of the tests.  Which means 80% of the time black affected nothing. 

I will do another set of tests on my fooseball table leg.  I will first try it blank on it's own black self, record the results then put a sheet of paper in front of it to see if it affects the results.  

Keep posting more info everyone! 

Keep the robotic dreams,
-BG
http://www.robocommunity.com/article/12811/Cool-Programs-To-Use-For-Roboquad-...

February 8, 2008 3:11 PM

As I stated, the material the surface is made from can play a more important role than the color variation.

Try different materials in your tests, and post the resuls.

Black felt. Black fleece. Black cotton. Black leather. Black plastic.

If you had night vision goggles or a camcorder that relied on infrared rather than light amplification, then you might be able to directly observe how IR reflective the different materials are. 

Alternatively, in other threads we've shown that a digital camera or camcorder is usually sensitive to the IR signals. They look like visible lights to the CCD recording the scene. You might be able to use a remote as an IR flashlight, and use a camera torecord how much of it reflects off of different surfaces. 

ScottE -- Member (always) & Moderator (when needed)

February 8, 2008 4:26 PM

I don't have all of those things however.  I do have black leather (many old and my current wallat).  I may have some black felt at hand.  Black Plastic could most likely find.

I must include that the two surfaces that caused him trouble were leather and well I don't know what the fooseball table leg is.  I guess it is painted wood. 

Keep the robotic dreams,
-BG
http://www.robocommunity.com/article/12811/Cool-Programs-To-Use-For-Roboquad-...

February 13, 2008 6:59 PM

I forgot to do the experiment!  I have been busy and I keep forgeting!  Don't worry I'll get to it next time I can say maybe friday the 15th (this month)?

Keep the robotic dreams,
-BG
http://www.robocommunity.com/article/12811/Cool-Programs-To-Use-For-Roboquad-...

February 21, 2008 1:18 PM

hi

just thought i would remind you that human's don't see true color's.

which is darker A or B ?

they are the same ........but robot's dont have this problem ..... or do they ?

i still believe

February 21, 2008 6:22 PM

Its not so much that we don't see true colour as it is that the brain plays tricks on us.

A and B are the same colour, because of the lighter and darker boxes around them, and that B is in "shadow", our brain perceives B as lighter. I find that if I stare at A for a few seconds without moving my eyes B appears to change colour (at least till my eyes move).

RoboGuide - Your guide to hacking all things WowWee

February 27, 2008 8:47 AM

Its true our brain gives our sence of light and dark and interprets what red should be refracted as ect and due to the way in which our eyes and brain communicate we dont actually see true to life colour/bright dark but to us thats normal........you should see how poor old rsv2 sees the world through his rgb value cmos sensors .......thats the thing there is many different ways of blending palettes or wavelengths, just what method do we see with? Is it subtractive? If so our vision is not true to life as the wavelengths may be absorbed by our retinas with external influences/Additives at play, it would leave room for errors in our vision, but just say we had additive vision we could more or less see in the dark we would be able to accurately asses the spectrum of the wavelength and then discern what light wavelength was reflecting back as our eyes would be sending light and analysing its spectrum that is returned off the reflectors so our eyes would be sorta like roboquads deep ir sacnning in additive, only issues there is the inherent black walls that we would walk into all the time lol 

February 27, 2008 10:11 AM

Welcome to the community, Tinkerer. Excellent screen name. It sums up what a lot of the community members enjoy about these robots.

I'm having just a wee bit of trouble decoding some of the content in your reply, but picked out a couple of things I thought I could comment on.

I don't think you refer to vision as subtractive or additive. That's for how we characterize how frequencies add or subtract before they reach the light sensors. The sensors, whether biological or technological, are sensitive to a range of frequencies. I guess by that logic the model is one of frequency absorbtion by the sensors, which is the basis for subtractive color mixing.

If you don't have multiple sensors that are sensitive to different frequency ranges, then you don't see any different colors. Following that logic, we don't add or subtract frequencies so much as we detect relative amounts of different frequencies. Our brain is doing some frequency mixing to provide a tally of

Our human color perception certainly varies from person to person, and can change within a single person based on medications or illness. I'm sure we're all aware of people who are color blind. They cannot differentiate between some groups of frequencies. Lots of background information on the mechanisms and extent of color perception deficiencies is available on the site listed below.

http://colorvisiontesting.com/ 

I've often pondered whether we really see anything the same as somebody else does. We're not born with the word "blue" being linked to a given neural input. As children we are taught to recognize and use labels for colors based on somebody else's perception of them.  

"What color is this?  This is red. Can you find something else red? No, that's not red, that's green."

That little scenario doesn't play out too well for children with the inability to differentiate between red and green. 

Our poor RSV2 has the crudest level of color detection. It can tell when something is mostly green, or mostly red. (the bowling pins and bowling ball) It's easily confused by the spectrum of the visible light used to shine on the objects.  Flourescent light versus incandescent light versus sunlight can confuse the color detection. It tries to detect a blend of colors we associate with skin. A manilla folder is just as likely to be detected as "human" as a flesh and blood person is. 

ScottE -- Member (always) & Moderator (when needed)

February 27, 2008 10:35 AM

Nice :) yer i was more or less trying to point out how if our vision systems did work in those manners how greatly it would affect what we see and our perception of it just so ppl could relate to themselfs as to why we dont hit black walls and robots do but i got off the beaten trail a bit lol .... thankyou for your most valuable information its helped me understand things that lil bit better.

I did notice though on the subject of rsv2 if u convert his his vision to cmyk colour format you get a lot higher accuracy on colour detection and speeds his vision systems up to around 12 fps versus a sad average of 6fps or so with rgb values 

February 27, 2008 11:38 AM

How do you do Mr. The Tinkerer, sorry I didn't get a welcome message to you I'll try today!!(:^))

"Life is not measured by the breaths we take, but by the moments that take our breath away."

February 27, 2008 3:07 PM

Coming back in here!

Ok my experiments using the foose ball table leg proved he took much longer to recognize the color black then when I put a big blue container lid in front of the leg.  100% of the tests showed this result. 

Keep the robotic dreams,
-BG
http://www.robocommunity.com/article/12811/Cool-Programs-To-Use-For-Roboquad-...

February 28, 2008 5:25 PM

so animated gifs wont upload?

February 28, 2008 7:53 PM

wait what?  I don't know what you are talking about.

Keep the robotic dreams,
-BG
http://www.robocommunity.com/article/12811/Cool-Programs-To-Use-For-Roboquad-...

February 28, 2008 10:22 PM

I,robot114 said: How do you do Mr. The Tinkerer, sorry I didn't get a welcome message to you I'll try today!!(:^))

Ok thankyou :) and its nice to meet you, im looking forwards to a very long stay with the robocommunity its a lifelong passion of mine robotics and the likes.

Im a I.T Professional by trade but yes robotics is really what i like the most computers just play a vital role for me :)

February 28, 2008 10:25 PM

Oh and its not just black that affects him ....due to the way i.r works i have found many matte paints and and soft surfaces wont detect well ....and in the oppsite swing of things highly reflctive things eg mirrors mess with him really bad.........

 try putting rsv2 in rsv2 to rsv2 mode and put him infront of a mirror and sit bak amazed!!!!!!

Roboquad goes nuts, roboreptile goes a bit weird too.

February 28, 2008 11:20 PM

Good day ladies and gentlemen(PC and all that . . . . )
Regarding RSV2, i have a specimen that's as blind as a bat(yes i know bats have and incredible sensory perception that runs rings round our own, but it expresses my point). You are ALL obviously a lot more clued up on all these matters than i, so can anyone suggest a good solution to my dilemma?

i've tried all of the vision settings, but so far i've had RSV2's recognition of the green ball once, that of the red skittle and once that of my hand once and all at the expense of my time nursing and educating the brat with little success and as for seeing a skittle then following it with the vision sensors and proceeding to perform a strike with the ball, i can honestly say, "it'll never happen!"

i've tried increasing the ambient light to the extent that only the floodlighting on a football pitch would be an improvement and this after adding a light source two each shoulder and a back pack to power them.

Maybe i just got a dud, but if there are any remedies or hacks that can be explained in laymans terms i would really appreciate your help.

Tinker. is that swap from rgb to cmyk something i missed in the manual? Must be a hack right?

You think you've got problems?

February 29, 2008 1:15 AM

There is a fairly simple and (relatively) common hack to add LED's to your V2's head that activate when the camera switches on, to aid in colour identification. Some people who did not want to violate their waranties have acheived similar improvements by mounting small torches to the head.

You may also want to try wiping the camera lense with a clean cloth.

RoboGuide - Your guide to hacking all things WowWee

February 29, 2008 6:57 AM

Thanx for the info Nocturnal, it's a nice neat little mod and if it works for my RSV2 it'll be a real nice touch. Looks good too!

You think you've got problems?

February 29, 2008 7:23 AM

Gotta love the differences in vocabulary and definition. It's all "english", but hereabouts a small torch would involve strapping  a flaming piece of wood to your robot's head.

I'm really enjoying the mental image of my RSV2 tottering about with flaming torches strapped to his head. 

That would undoubtably be enteraining until the torch burnt down to the robot's head (and for some short time after that).

ScottE -- Member (always) & Moderator (when needed)

Discussion:    Add a Comment | Back to Top | Comments 1-25 of 41 | Latest Comment | 1 2 Next »

Add Your Reply

(will not be displayed)

Email me when comments are added to this thread

 
 

Please log in or register to participate in this community!

Log In

Remember

Not a member? Sign up!

Did you forget your password?

You can also log in using OpenID.

close this window
close this window