DAIHATSU Indonesia Masters 2021 Nov 16 to Nov 21 WT Super 750

Discussion in '2021 Tournaments' started by samkool, Nov 15, 2021.

  1. durmogati

    durmogati New Member

    Joined:
    Nov 4, 2021
    Messages:
    9
    Likes Received:
    0
    Location:
    astino
    video replay is better than eagle eye. more difficult to manipulate
     
  2. samkool

    samkool Regular Member

    Joined:
    Oct 31, 2004
    Messages:
    2,497
    Likes Received:
    1,675
    Occupation:
    too pre-occupied to work
    Location:
    the next world tour tournament
    absolutely not. i've always been ambivalent as to the 'shadow' used to represent the shuttlecock (and tennis ball/volleyball in those sports), but more on that below.
    obviously not at 100%. one or more of the 8 cameras used to triangulate the shuttlecock was not working.
    that's outright flat-earther talk. read up on how hawkeye works and who is operating it before you respond with such conspiratorial nonsense.
    yes, now you are making sense and this is a bwf problem. hawkeye is not the final decision, but it appears to be because the umpires agree with it 100% of the time. the umpire can overrule hawkeye but when have we ever known a bwf umpire to overrule anyone or anything for any reason?

    my ambivalence with hawkeye:

    since balls are round does it take into account the compression of the ball in tennis & volleyball? the ball hits the lines with varying amounts of force depending on the shot. part of the ball will be on top of/hovering over the line without actually touching the line. is hawkeye coded to assume how much surface of a ball is actually touching the ground given the force and trajectory driving it? since a shuttlecock cork doesn't compress when it hits the ground is hawkeye coded to assume only roughly 2-3 millimetres of the cork are actually touching the ground? a regulation line is 40 mm wide. the circular shadow hawkeye displays for a shuttlecock should be way way smaller in diameter than what we see relative to the width of the line they show.

    we have accepted hawkeye results because human emotions have been removed. but it is not foolproof yet for many reasons.
     
  3. Ballschubser

    Ballschubser Regular Member

    Joined:
    May 17, 2019
    Messages:
    339
    Likes Received:
    255
    Location:
    Germany
    There will be always technical limitations. If some cameras only have an obscured view of the shuttle, net caughts, drift etc. will make it hard to determine the excat impact position. I think, that a long trajectory will be quite accurate, whereas a short obscured trajectory will be quite hard even for an automatic system. Especially if the receiver will move into and over the shuttle, the view could be hard for the camera and each camera has only a 2D view of the scene, so you need several camera views to determine a 3D trajectory of the shuttle, hoping that it will not tumble or deflected by the net...
     
  4. Nine Tailed Fox

    Nine Tailed Fox Regular Member

    Joined:
    May 10, 2015
    Messages:
    10,031
    Likes Received:
    2,401
    Location:
    New Delhi
  5. j4ckie

    j4ckie Regular Member

    Joined:
    Aug 12, 2010
    Messages:
    6,305
    Likes Received:
    1,571
    Location:
    Germany
    Hawkeye should be based on video tracking, so it should (in theory) show the ball exactly as it did contact the floor. While it can be inaccurate (cameras can be obscured, and the lighting is inconsistent and way worse than in any industrial computer vision application), it's still far more accurate than humans, especially with fast-moving objects (e.g. shuttle moving quickly across the line on flat drives).
    If it can one day be scaled to that it automatically judges every single shot, that'd be great, but I guess it'd not really feasible to be set up on all courts..
     

Share This Page