Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Sensors implimentation

  1. May 4, 2010 #1
    Hi, I am a final year student whos doing engineering degree. My final year project involves tracking the speed of a swimmer inside a pool.
    The current problem is that i dont have a good way of monitoring the swimmer's speed, which needs to be quite accurate(centimeters). Due to different stroke patterns and unregular shape of human body, it is very hard to find a way of getting the swimmer's speed during motion.

    I've been thinking about infra-red sensors, ultra sonic sensors and also other methods.
    Is there any way of using sensors wisely and place them the right way so that i can accurately get the motion of the C.O.M. of the swimmer OR the mid-point of the body length?

    Very appreciated for any suggestions or ideas.

    Regards, Kent Zhou.
     
  2. jcsd
  3. May 4, 2010 #2
    An overhead camera looking down on the swimmer measuring a target on the swimmer such as line or cross on the swimmer's back?
     
  4. May 4, 2010 #3

    berkeman

    User Avatar

    Staff: Mentor

    Two comments.

    First, since this is for a graded school project, you need to do the bulk of the work. Do not try to have the PF do your project for you. Even though this post probably qualifies under the "project" rule to be posted in the general technical forums like EE, you still need to show some effort on your part. We do not do your schoolwork for you.

    Second, this is a VERY HARD project! Holy smokes. I'm a competitive swimmer, and I'm hard pressed to figure out a way to do it. Even with the video suggestion put forth so far, you are not going to see your sensor/reflector/marker much of the time from above. And most competitive pools are not deep enough for a single camera to sense the position of the swimmer with an under-Speedo-side marker.

    What are YOUR thoughts so far? Are you a swimmer? Have you watched some competitivbe swimming videos on the Net? Let's hear your ideas please before this thread goes much farther.
     
  5. May 4, 2010 #4
    Thank you for replying. What you have said above needs image processing and it is hard. what i have forgot to say is that the project is supposed to be "low cost" which limits how much resources we can have/spend on this project.

    our design is actually on the side of the pool, plus the system needs to be portable. so....cant really bolt anything or suspend anything.

    Thank you for your ideas anyway.
     
  6. May 4, 2010 #5

    Thank you very much for your time. I do agree with you that I need to do most of the work which I am trying to get a solution right now. But sometimes, the views from experts could widen my ideas and for which i can get my work done in a relatively better quality, also saves time. XD.
    Just some ideas needed or some devices thats available out there somewhere which I could not find with google.

    Hmmm, I've had look at similar devices used in the olympics. They are really complicated and bulky and as I've replied to skeptic2 that this project has limited resources which can be spend on.

    Theres a platform that runs along side of the pool which carries the cameras. The main ideas would be to have sensors placed on the platform and maybe some on the swimmer, possibly a band which can be picked up by the camera and then some image processing operating real time. TOO EXPENSIVE. What I have in mind is to have a own built PID controller or other types of controller (PCB) to control the speed of the platform.
    Without too much expensive real time programming capability = can't use image processing, or anything like that.

    Well, I am not a swimmer myself but i certainly can swim. I have talked to coaches about different strokes and how they will move, etc. Different strokes can have quite a different velocity pattern and the CoM moves up and down as well. Unlike a car or anything solid, theres an edge or fixed point which can be tracked..... This is really bugging me....

    My idea now is to have array of sensors lined up horizontally and have a little algorithm which determines wheres the center of the body for different strokes. and honstly, cant find any common patterns of arms/legs for different strokes....

    Looking forward for your reply.

    Thank you, Kent Zhou
     
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook