Electronic Projects, forums and more.

Go Back   Electronic Circuits Projects Diagrams Free > Electronics Categories > Robotics Chat


Robotics Chat Specific to discussions about robots and the making of.

Reply
 
Tools
Old 30th July 2006, 06:28 PM   #1
Default outdoor IR object detection

Hi all!

Im building robot for outdoor competition (like US DARPA grand challenge but smaller) and i want to use IR to object detection. I dont need information about how far object is, i only need detect object is/isnt before robot. I would like to detect object in cca. 2 meters. Can you give me some tips to artiles or some materials about this? I want use IR, not RADAR or camera or something else. thx
SniperCZE is offline  
Old 30th July 2006, 06:38 PM   #2
Default

2m is only possible with focused IR- which means optics- which means hard to DIY. Look for one of the sharp IR sensors. Perhaps you can find a proximity version and stick it on a rotating mast or something or if not, just use one of the rangefinder versions, and then you can just ignore all readings greater than 2m.

The beamwidth is not as wide as you want, nor as narrow as you might think. At a distance of 2m you might only need 3 or 4 sensors in an array to get a 180 degree spread (although as objects get closer, gaps form between the beams and the object is more likely to "dissapear").

Last edited by dknguyen; 31st July 2006 at 06:11 AM.
dknguyen is offline  
Old 6th August 2006, 10:23 AM   #3
Default

with a normal IR and 38 k modulation , I have managed only 4 feet
without any false triggering
IR reflective power loss is very high . I doubt if you can get more distance.
Ultrasonic echo would be the only choice.
__________________
K.S.Sankar
mostek is offline  
Old 6th August 2006, 12:21 PM   #4
Default

use Sonar for accurate distance measurement
sometimes 38KhzIR+TSOP will be helpful for good distance in meters

IRLED+phtodiode gives upto 5cm.
So choose any of these
http://booksbybibin.14.forumer.com/viewtopic.php?t=115

Bibin John
www.booksbybibin.tk
bibinjohn is offline  
Old 6th August 2006, 03:43 PM   #5
Default

I know some manufactures who produce these IR sensor .It is used for detecting thief who break your home .The distance is 2,3m .
If you want ,why don't you buy them to discover the schematic
i think it is good .
For me ,I use a TSOP to receiver IR from TV remote ,and the MCU to decode the signal from it.the software is very complicated because of the interference from the other IR source
__________________
E-DESIGN prestigious and quality
http://pg.photos.yahoo.com/ph/giaosu...765scd&.src=ph
contact me:giaosucan2000@yahoo.com
giaosucan is offline  
Old 15th August 2006, 12:55 AM   #6
Default

So, I take a decision to use ultrasound detector instead of IR. I think it will be the best thing for me
SniperCZE is offline  
Old 15th August 2006, 02:30 PM   #7
Default

parrallax has a product that is good for 5 meters and is a motion detector; this coupled with stereo sonar and stereo IR should help your robot pinpoint a moving target. Then your robot would have three sensor ranges, close, medium and long. The Sonar would provide detection up to 6 - 12 ft with a 1 ft accuracy, and Oft difference below 6ft. Also, it is important to remember that IR in an outdoor enviroment is not allways accurate, as dust and debree will cause miss readings, so having two IRs will help determine if the target is near and within range of attacking.
Aphromann is offline  
Old 15th August 2006, 02:38 PM   #8
Default

Quote:
Originally Posted by SniperCZE
So, I take a decision to use ultrasound detector instead of IR. I think it will be the best thing for me
Probably. Just remember sonar is less accurate (and doesn't work as far) if there is wind. Except it will still work (whereas IR may not work at all). YOu can also tell if there is wind around to know your sonar won't work as well. With IR, you cannot really tell.
dknguyen is offline  
Old 7th September 2006, 01:16 PM   #9
Lightbulb well i think we can tell....

Quote:
Originally Posted by dknguyen
With IR, you cannot really tell.
Well there is a simple technique to know if the IR you get is wrong or not..

using a micro controller, you have to establish 2 readings, store them, and compare them,
1- send IR, and read the sensor output, store it
2- Stop sending IR,and read the sensor output, store it

if you get (1) from the sensor in both cases, then the IR is not working good, dure to surrounding light, noize, etc...

if you get (1) then (0), there is an abstacle detected.

if you get (0) in both cases, there is nothing infront of you, for SURE!

think about it..
__________________
Ibrahim Kamal
check my electronics and robotics page: http://www.ikalogic.com/
ikalogic is offline  
Old 7th September 2006, 01:44 PM   #10
Default

Quote:
Originally Posted by ikalogic
Well there is a simple technique to know if the IR you get is wrong or not..

using a micro controller, you have to establish 2 readings, store them, and compare them,
1- send IR, and read the sensor output, store it
2- Stop sending IR,and read the sensor output, store it

if you get (1) from the sensor in both cases, then the IR is not working good, dure to surrounding light, noize, etc...

if you get (1) then (0), there is an abstacle detected.

if you get (0) in both cases, there is nothing infront of you, for SURE!

think about it..
How would you know if something had moved out of range or under range (only applies while moving)? It's impractical to constantly stop to check some it's mechanically demanding (and time consuming especially for a contest) to start and stop so often since that interference could kick in at any time.

Last edited by dknguyen; 7th September 2006 at 01:50 PM.
dknguyen is offline  
Old 7th September 2006, 02:13 PM   #11
Default i've tested it..

Quote:
Originally Posted by dknguyen
How would you know if something had moved out of range or under range (only applies while moving)? It's impractical to constantly stop to check some it's mechanically demanding (and time consuming especially for a contest) to start and stop so often since that interference could kick in at any time.

Well honnestly i tested it, and it worked perfectly, using a microcontroller at only 12Mhz, you can switch it on and off very fast, faster than any moving object around.

but about knowing if something had moved out of range or under range, i guess you are right.. it may lack this point.
__________________
Ibrahim Kamal
check my electronics and robotics page: http://www.ikalogic.com/
ikalogic is offline  
Old 7th September 2006, 03:46 PM   #12
Default

Ah, yeah I guess it can switch on and off pretty fast.
dknguyen is offline  
Old 12th September 2006, 08:28 AM   #13
Default

use a 38khz IR module. it will give you UP TO 3m with out any false result.
http://robotroom.com/Infrared555.html
http://robotroom.com/InfraredTransmitter.html
__________________
Tabriz
arash_ele is offline  
Old 12th September 2006, 09:15 AM   #14
Default

Quote:
Originally Posted by ikalogic
Well honnestly i tested it, and it worked perfectly, using a microcontroller at only 12Mhz,
"ONLY" 12MHz, that's a pretty fast clock speed!.

Microcontrollers run many, many times faster than the robot itself, processor speed isn't an issue at all - the programme will spend almost all it's time in loops waiting for the hardware!.
__________________
PIC programmer software, and PIC Tutorials at:
http://www.winpicprog.co.uk
Nigel Goodwin is offline  
Old 2nd October 2006, 09:42 AM   #15
Default

I take a decision to use ultrasound detector instead of IR . I think it will be the best thing for me
izcidogan is offline  
Reply

Tags
detection, object, outdoor

Thread Tools
Display Modes




All times are GMT. The time now is 03:20 AM.


Electronic Circuits  |  Learning Electronics
eXTReMe Tracker