pith. sign in

arxiv: 2605.28011 · v1 · pith:AM75TORHnew · submitted 2026-05-27 · 💻 cs.CV

Automated Estimation of Impact Time, Impact Location, and Shuttlecock Speed in Badminton Smashes Using Event Cameras

Yudai Washida , Yuto Kase , Kai Ishibe , Ryoma Yasuda , Sakiko Hashimoto This is my paper

Pith reviewed 2026-06-29 13:22 UTC · model grok-4.3

classification 💻 cs.CV
keywords event camerasbadminton smashimpact time estimationimpact locationshuttlecock speedtrajectory inflectionellipse fittingBland-Altman validation
0
0 comments X

The pith

Two synchronized event cameras automatically estimate impact time, location on the racket, and shuttlecock speed in badminton smashes.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper develops an integrated measurement approach that replaces conventional high-speed camera systems, which trade off temporal resolution against data volume and setup complexity. Swing intervals are identified from event rate statistics, impact time is read from an inflection in the lateral-view shuttlecock trajectory, impact location is found by fitting an ellipse to the racket face in the rear view, and post-impact speed is computed in the sagittal plane. On 124 smash trials the estimates show small average differences from a high-speed reference with no proportional bias, and the method succeeds for time and speed in every trial and for location in 93.5 percent of trials.

Core claim

The method using two synchronized event cameras estimates impact time from shuttlecock trajectory inflection in the lateral view, determines impact location by ellipse fitting to the racket face in the rear view, and calculates shuttlecock speed in the sagittal plane, yielding biases of 1.84 ms for time, 3.45 mm and -1.92 mm for the two location axes, and -1.00 m/s for speed against a high-speed camera reference across 124 trials, with no proportional bias observed for any quantity.

What carries the argument

Two synchronized event cameras that supply event-rate swing detection, lateral trajectory inflection for timing, rear-view ellipse fitting for location, and sagittal-plane velocity calculation.

If this is right

  • Impact time, location, and speed can be obtained automatically within a single trial for combined performance and equipment assessment.
  • The estimates exhibit only small biases and no proportional error relative to high-speed camera references.
  • The approach succeeds for time and speed in all analyzable trials and for location in 93.5 percent of them.
  • Preparation and data-handling demands are lower than those of conventional high-resolution systems.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The same camera pair could support repeated testing on a standard court without moving heavy reference equipment.
  • Event-stream processing might be extended to give immediate numerical feedback on smash quality during a training session.
  • The inflection-plus-ellipse pipeline could be tested on other racket sports that produce brief, high-speed contacts.

Load-bearing premise

The lateral event stream must contain a detectable trajectory inflection exactly at racket-shuttlecock contact, and the rear-view image must permit accurate ellipse fitting to the racket face without occlusion or motion blur.

What would settle it

A fresh collection of smash trials in which the estimated impact times fall outside the reported 95 percent confidence intervals of the high-speed reference or in which ellipse fitting fails on more than a small fraction of trials would show the estimates are not reliable.

read the original abstract

Quantifying impact phenomena in badminton smashes is important for evaluating both athletic performance and equipment; however, conventional measurement systems involve trade-offs between temporal resolution, data efficiency, and preparation effort. This study proposes a measurement method using two synchronized event cameras to automatically estimate impact time, impact location on the racket face, and post-impact shuttlecock speed in an integrated manner within the same trial. The swing interval was detected from event rate statistics, impact time was estimated from the shuttlecock trajectory inflection in the lateral-view event data, impact location was determined by ellipse fitting to the racket face in the rear-view event image, and shuttlecock speed was calculated in the sagittal plane. To validate the proposed method, Bland-Altman analysis was performed against a high-speed camera-based reference method using 125 smash trials from five players. Impact time and shuttlecock speed were estimated in all 124 analyzable trials, and impact location was estimated in 93.5% (116/124). The bias (95% CI) for impact time, medio-lateral impact location, longitudinal impact location, and shuttlecock speed were 1.84 ms (1.45 to 2.23), 3.45 mm (2.18 to 4.72), -1.92 mm (-2.97 to -0.88), and -1.00 m/s (-2.46 to 0.46), respectively. No proportional bias was observed for any metric. These results suggest that the proposed method can serve as a useful tool for integrated assessment of badminton smash performance and equipment in practical settings.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 2 minor

Summary. The paper claims that two synchronized event cameras enable automatic, integrated estimation of impact time (from lateral-view trajectory inflection), impact location (via ellipse fitting on the rear-view racket face), and post-impact shuttlecock speed (in the sagittal plane) for badminton smashes. On 124 analyzable trials from five players, Bland-Altman analysis against a high-speed camera reference reports small biases without proportional error (1.84 ms for time, 3.45 mm and -1.92 mm for medio-lateral/longitudinal location, -1.00 m/s for speed), with 100% success for time/speed and 93.5% for location.

Significance. If the core assumptions hold, the method could provide a practical, low-preparation tool for simultaneous performance and equipment assessment, leveraging event cameras' high temporal resolution and data efficiency over conventional systems. The multi-player dataset and statistical validation approach support claims of utility in realistic settings.

major comments (2)
  1. [Impact time estimation] Impact time estimation (abstract and methods): The method asserts that the inflection point in the lateral-view reconstructed trajectory marks the exact racket-shuttlecock contact instant, yet this is supported solely by agreement (bias 1.84 ms) with the high-speed camera reference on the same 124 trials. No independent ground-truth sensor (force plate, microphone, or strain gauge) is used, so any shared systematic timing offset between the two optical approaches would remain undetected.
  2. [Impact location determination] Impact location determination (abstract): The 93.5% success rate (116/124 trials) for ellipse fitting indicates that rear-view event images are not always free of artifacts, but the manuscript does not characterize the failure modes (occlusion, motion blur, or synchronization issues) or analyze their distribution across players/trials, which directly affects reliability claims for practical use.
minor comments (2)
  1. The abstract states 125 trials were collected but only 124 were analyzable; explicit criteria for exclusion and whether any ellipse-fitting parameters were tuned on the validation set should be added to allow assessment of potential overfitting or selection bias.
  2. Details on camera synchronization error propagation and how it affects the reported 1.84 ms bias should be quantified, as small timing offsets could influence the no-proportional-bias conclusion.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive feedback on our manuscript. The comments highlight important aspects of validation and reliability that we address below. We provide point-by-point responses to the major comments.

read point-by-point responses

  1. Referee: [Impact time estimation] Impact time estimation (abstract and methods): The method asserts that the inflection point in the lateral-view reconstructed trajectory marks the exact racket-shuttlecock contact instant, yet this is supported solely by agreement (bias 1.84 ms) with the high-speed camera reference on the same 124 trials. No independent ground-truth sensor (force plate, microphone, or strain gauge) is used, so any shared systematic timing offset between the two optical approaches would remain undetected.

    Authors: We agree that an independent sensor (e.g., force plate or microphone) would strengthen the validation by ruling out potential shared optical timing offsets. The high-speed camera remains a standard reference in sports biomechanics literature for kinematic validation. The small bias (1.84 ms) and absence of proportional error provide evidence of agreement within the limits of optical methods. In the revised manuscript, we will add an explicit discussion of this limitation, noting the reliance on the high-speed camera reference and the potential for undetected systematic offsets. revision: partial

  2. Referee: [Impact location determination] Impact location determination (abstract): The 93.5% success rate (116/124 trials) for ellipse fitting indicates that rear-view event images are not always free of artifacts, but the manuscript does not characterize the failure modes (occlusion, motion blur, or synchronization issues) or analyze their distribution across players/trials, which directly affects reliability claims for practical use.

    Authors: We concur that characterizing the failure modes is necessary to support reliability claims. The eight unsuccessful trials will be analyzed in the revision, with descriptions of observed artifacts (primarily occlusion by the player's arm or body, and occasional motion-related event sparsity) and their distribution across the five players. This analysis will be incorporated into the Results or Discussion section of the revised manuscript. revision: yes

Circularity Check

0 steps flagged

No significant circularity; validation uses external reference

full rationale

The paper describes an algorithmic pipeline (event-rate swing detection, trajectory-inflection timing, ellipse fitting on rear-view events, sagittal-plane speed calculation) whose outputs are compared to an independent high-speed-camera reference via Bland-Altman analysis on 124 trials. No equation or parameter is shown to be algebraically equivalent to the reported bias values by construction, no fitted input is relabeled as a prediction, and no self-citation chain is invoked to justify the core premises. The method therefore remains self-contained against external benchmarks.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 0 invented entities

The central claim rests on the assumption that event-camera data streams contain a detectable inflection and an unambiguous racket ellipse at contact; no free parameters are explicitly fitted inside the estimation equations, and no new physical entities are postulated.

axioms (2)
  • domain assumption Event cameras provide sufficient spatial and temporal resolution to capture shuttlecock trajectory inflection and racket-face ellipse at the moment of impact.
    Invoked in the impact-time and impact-location estimation steps; treated as given rather than derived.
  • domain assumption The two cameras remain rigidly synchronized and calibrated throughout each trial.
    Required for combining lateral and rear views into a single 3-D speed calculation.

pith-pipeline@v0.9.1-grok · 5844 in / 1511 out tokens · 24208 ms · 2026-06-29T13:22:35.646701+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

14 extracted references · 12 canonical work pages

  1. [1]

    Sci Rep 13:14060

    Towler H, Mitchell SR, King MA (2023) Effects of racket moment of inertia on racket head speed, impact location and shuttlecock speed during the badminton smash. Sci Rep 13:14060. https://doi.org/10.1038/s41598-023-37108-x

    work page doi:10.1038/s41598-023-37108-x 2023

  2. [2]

    J Sports Sci 38(21):2471–2478

    McErlain-Naylor SA, Towler H, Afzal IA, Felton PJ, Hiley MJ, King MA (2020) Effect of racket-shuttlecock impact location on shot outcome for badminton smashes by elite players. J Sports Sci 38(21):2471–2478. https://doi.org/10.1080/02640414.2020.1792132

    work page doi:10.1080/02640414.2020.1792132 2020

  3. [3]

    Appl Sci 15(2):844

    Ramasamy Y , Yeap MW, Towler H, King M (2025) Intra-individual variation in the jump smash for elite Malaysian male badminton players. Appl Sci 15(2):844. https://doi.org/10.3390/app15020844

    work page doi:10.3390/app15020844 2025

  4. [4]

    New J Phys 17:063001

    Cohen C, Darbois Texier B, Quéré D, Clanet C (2015) The physics of badminton. New J Phys 17:063001. https://doi.org/10.1088/1367-2630/17/6/063001

    work page doi:10.1088/1367-2630/17/6/063001 2015

  5. [5]

    Phys Scr 101:125007

    Collet E (2026) Shuttlecock velocity decay after smash and slice shots in badminton. Phys Scr 101:125007. https://doi.org/10.1088/1402-4896/ae5361

    work page doi:10.1088/1402-4896/ae5361 2026

  6. [6]

    IEEE Trans

    Gallego G, Delbruck T, Orchard G, Bartolozzi C, Taba B, Censi A et al (2022) Event- based vision: a survey. IEEE Trans Pattern Anal Mach Intell 44(1):154–180. https://doi.org/10.1109/TPAMI.2020.3008413

    work page doi:10.1109/tpami.2020.3008413 2022

  7. [7]

    In: 2024 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW)

    Gossard T, Krismer J, Ziegler A, Tebbe J, Zell A (2024) Table tennis ball spin estimation with an event camera. In: IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW), pp 3347–3356. https://doi.org/10.1109/CVPRW63382.2024.00339

    work page doi:10.1109/cvprw63382.2024.00339 2024

  8. [8]

    In: 2024 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW)

    Nakabayashi T, Higa K, Yamaguchi M, Fujiwara R, Saito H (2024) Event-based ball spin estimation in sports. In: IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW), pp 3367–3375. https://doi.org/10.1109/CVPRW63382.2024.00341

    work page doi:10.1109/cvprw63382.2024.00341 2024

  9. [9]

    In: Sports Analytics (ISACE 2025)

    Kase Y , Ishibe K, Yasuda R, Washida Y , Hashimoto S (2025) Locating tennis ball impact on the racket in real time using an event camera. In: Sports Analytics (ISACE 2025). Lecture Notes in Computer Science, vol 15925. Springer, Cham, pp 99–115. https://doi.org/10.1007/978-3-032-06167-6_8

    work page doi:10.1007/978-3-032-06167-6_8 2025

  10. [10]

    Sports Informatics and Technology 2025, Session ID B-3-1

    Yasuda R, Hashimoto S, Kase Y , Ishibe K, Washida Y (2025) Estimating shuttlecock speed in badminton smashes with an event camera. Sports Informatics and Technology 2025, Session ID B-3-1. https://doi.org/10.82713/sit.2025.0_B-3-1 (in Japanese)

    work page doi:10.82713/sit.2025.0_b-3-1 2025

  11. [11]

    Biochem Med (Zagreb) 25(2):141–151

    Giavarina D (2015) Understanding Bland Altman analysis. Biochem Med (Zagreb) 25(2):141–151. https://doi.org/10.11613/BM.2015.015

    work page doi:10.11613/bm.2015.015 2015

  12. [12]

    Stat Methods Med Res 22(6):630–

    Zou GY (2013) Confidence interval estimation for the Bland–Altman limits of agreement with multiple observations per individual. Stat Methods Med Res 22(6):630–

    2013

  13. [13]

    https://doi.org/10.1177/0962280211402548

    work page doi:10.1177/0962280211402548

  14. [14]

    Statutes: Section 4.1 – Laws of Badminton

    Badminton World Federation (BWF). Statutes: Section 4.1 – Laws of Badminton. Available at: https://corporate.bwfbadminton.com/statutes/#1513733461252-a16ae05d- 1fc9(Accessed: 18 May 2026). Fig. 1 Overview of the proposed system. (a) Experimental setup in an indoor gymnasium, showing sensor placement and the measurement environment. (b) Processing pipeline...

    2026