Shop OBEX P1 Docs P2 Docs Learn Events
trying to build a heat tracker using IR devices to track a humans body heat — Parallax Forums

trying to build a heat tracker using IR devices to track a humans body heat

mikedivmikediv Posts: 825
edited 2009-10-06 21:10 in General Discussion
Hi guys I am trying to build a device that will track (by heat) a person when they walk into range of my device. I have been trying to use my BoeBot with the Parallax IR devices that come with BoeBot but it seems that they are not anywhere near sensitive enough
I can get it to detect and kind of track a lit match but it has to be about 12 inches near the IR's does anyone have any thoughts or suggestions is there a more sensitive device I should be using .. What I would really like to do is make a small parabolic dish and put the IR in the middle and see if that would increase sensitivity??????

Comments

  • TimmooreTimmoore Posts: 1,031
    edited 2009-10-02 01:28
    Take a look at the mlx90614 or the tpa81, both will detect humans though I dont remember the range. but it was more than 12"
  • ElectricAyeElectricAye Posts: 4,561
    edited 2009-10-02 03:37
    I've got some of these sensors in my house and I know they can run on backup batteries. In the night, they can detect a person inside my house from 50 feet away. I think they only cost about $20 for the whole unit.
    pdf
    pdf
    MotionSensor copy.pdf
    989K
    pdf
    pdf
    MotionSensor2 copy.pdf
    1019K
  • LeonLeon Posts: 7,620
    edited 2009-10-02 16:31
    I used to work for Irisys:

    www.irisys.co.uk

    One of our prototype people counters was fitted with a couple of servo motors and a gimbal system, so that it would actually track people as they moved around and a video camera would display what it was tracking. It actually homed in on the warmest part of the body it could find when one stood close to it - the groin area with males. It didn't have enough vertical movement to detect the head.

    Leon

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Amateur radio callsign: G1HSM
    Suzuki SV1000S motorcycle
  • mikedivmikediv Posts: 825
    edited 2009-10-03 00:36
    Thanks guys. Leon that's exactly the kind of thing I want to do. ElectricAye looking at the datasheet my burglar alarm has somehting similar to that but like what Leon posted I should have explained better, I have a turret controlled by servos X,Y,Z maybe that's not correct
    It can track left right up and down a full 180 degrees I have the servos being controlled for now my Basic Stamp OEM board only because I found some software I really want to use Prop chip but anyway I want to be able to have someone walk into the room and the device will track there movement there will be a video camera mounted on my tripod rig so while you are walking around the video camera will follow you "track you" I found a program for the BoeBot that had BoeBot following a flame around using some IR devices . It does work but its very limited I tried to expand on this or apply it to my project but it will not track anything colder than a match and even with that it has to be about 6 to 12 inches away form the IR sensors I even have given some thought to using the Ping but at the end of the day I want to be able to track a person at night by heat detection it doesn't have to be perfect its just for fun but I would like to be able;e to do it
  • ElectricAyeElectricAye Posts: 4,561
    edited 2009-10-03 01:35
    Mike,

    I have no idea how well these work, but this was in a folder I had from a year or so ago.

    Hope it helps.
    pdf
    pdf
    PIRpanasonic.pdf
    1M
  • TimmooreTimmoore Posts: 1,031
    edited 2009-10-03 01:53
    I have found PIR's tend to have a very wide view > 90degree, a tpa81 has 8 sensors each 5x6 degrees, mlx90614 you can get 10, 35 and 90 degree sensors. If you want to track which one you want depends on how close you need to track. PIRs aren't going to let you track very well. TPA81 will control a servo directly to track. I used 3 PIRs one time to get 360degree coverage, if I remember they were the sparkfun PIRs.
  • ZootZoot Posts: 2,227
    edited 2009-10-03 03:09
    Have you seen this? Does just what you want using the Devantech thermal array sensor:

    norrislabs.com/Projects/FollowMeRov/index.html

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    When the going gets weird, the weird turn pro. -- HST

    1uffakind.com/robots/povBitMapBuilder.php
    1uffakind.com/robots/resistorLadder.php
  • ElectricAyeElectricAye Posts: 4,561
    edited 2009-10-03 04:01
    Phil Pilgrim built this totally awesome device. I don't know if it can do body heat, but it might serve as an inspiration for your tracking system.

    I think this is just the "coolest" thing. lol.gif

    http://forums.parallax.com/showthread.php?p=753166
  • Carl HayesCarl Hayes Posts: 841
    edited 2009-10-04 21:44
    Include sun trackers in your research. You'll need greater sensitivity, but conceptually a device that steers a PV array toward the sun is the same sort of design problem as a device that steers a TV camera (or a rifle) toward a person.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    · -- Carl, nn5i@arrl.net
  • mikedivmikediv Posts: 825
    edited 2009-10-04 23:48
    Thanks guys, ElectricAye Phil does come up with some pretty cool stuff I did not even know Parallax had a new a new heat semsor just more stuff to buy, Zoot that is awesome I should have thought to check his site thanks guys for the help
    I love the idea of using the PIC 16f684 I just bought a few to build some other projects I had seen online ,, Wow Zoot did you see the cost for the TPA81 HOLY!! 110 bucks a piece and if I am reading the article right he used 8 of them???? Well I have to admit that puts that out of my reach I am positive my wife would divorce me not that its necessarily a bad thing but I guess I will keep trying .. ElectricAye I actually did give some thought to using the Parallax PIR I bought a couple but like TIM said they are very broad I could not get them down to focal point that was anything I could use . You would think this would be a very easy project even for a beginner the concept is not all that hard its just finding the right devices I guess
  • ZootZoot Posts: 2,227
    edited 2009-10-04 23:53
    No, Mr. Norris used ONE thermal array -- the thermal array sensor IS eight sensors configured as an array (hence the name). And yes it is pricey, which is why I have yet to purchase one smile.gif -- something I was going to do on the spot the first time I saw one of Steve's "follow me" 'bots.

    I mentioned because it really is ideal and pretty much ready-to-go with simple code. But a single thermal sensor on a servo might suffice as well and bring the cost down a bit.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    When the going gets weird, the weird turn pro. -- HST

    1uffakind.com/robots/povBitMapBuilder.php
    1uffakind.com/robots/resistorLadder.php
  • LilDiLilDi Posts: 229
    edited 2009-10-06 20:29
    After looking at some documentation on the TPA81 thermal array, it appears the effective range is about 6 feet. A PIR sensor with have a much greater range of about 25 feet.
  • ZootZoot Posts: 2,227
    edited 2009-10-06 21:10
    LilDi -- however the two are very different kinds of sensors -- the thermal arrays can be used for nearly real-time tracking and sensing of thermal differences and in a very narrow field of focus. The PIR has a very wide focus and requires settling of the space before detection, etc.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    When the going gets weird, the weird turn pro. -- HST

    1uffakind.com/robots/povBitMapBuilder.php
    1uffakind.com/robots/resistorLadder.php
Sign In or Register to comment.