Shop OBEX P1 Docs P2 Docs Learn Events
"resolution" of parallax ultrasonic sensors? — Parallax Forums

"resolution" of parallax ultrasonic sensors?

CuriousOneCuriousOne Posts: 931
edited 2014-02-15 08:30 in Accessories
Hello.

I want to do the following thing: take the ultrasonic sensors, mount them on a plate, which will be rotated around it's axis by servo. The idea is, to "scan" the area with sensors and find a way to go there. So the question is. Say, two persons are standing 4 feet away from sensor, with distance of 2 feet between them. Will sensor be able distinguish persons and space between them?

Comments

  • Chris SavageChris Savage Parallax Engineering Posts: 14,406
    edited 2014-02-11 16:24
    People aren't good objects for an ultrasonic sensor as a general rule because we're not very reflective. Assuming we do reflect the signal though, you would either detect or not. The sensor doesn't provide an indication of how many things it sees and the spaces between them. To get something like that you would have to rotate the sensor (using our PING))) mounting bracket) and take multiple reading, using that data to determine where objects are in front of you and the space between them. There is an application called, "PING)))Dar" which may help.

    http://forums.parallax.com/showthread.php/86110-Ping)))Dar-a-Radar-Style-Display
  • CuriousOneCuriousOne Posts: 931
    edited 2014-02-14 07:11
    Yes, I'm planing to mount a sensor on servo and rotate it. So anyone tried practical distance, from which sensor detects the human body?
  • ZootZoot Posts: 2,227
    edited 2014-02-14 13:35
    Unless the person is wearing a really, really "fuzzy" sweater, people will reflect ultrasonic frequencies from the Ping))) all right, esp. if you are "sweeping" for distances and aren't depending on one fixed angle for a reflextion.
  • CuriousOneCuriousOne Posts: 931
    edited 2014-02-15 01:55
    My idea is to develop a robot which can move trough crowd, without colliding with anyone.
  • ZootZoot Posts: 2,227
    edited 2014-02-15 07:42
    You'd probably be better off with a combo of IR and sonar to make sure you catch everything (and fast enough).
  • prof_brainoprof_braino Posts: 4,313
    edited 2014-02-15 08:30
    One thing to bear in mind is that all targets represent MULTIPLE distances (unless you are in the exact center of a hollow sphere).
    The reading one gets are typically a set of ranges, ranging from the furthest detectable point to the nearest detectablbe point.

    E have to get pretty fancy to dtermine which set of readings belong to different targets, when multiple targets may be presented.
Sign In or Register to comment.