首页 | 本学科首页   官方微博 | 高级检索  
相似文献
 共查询到20条相似文献，搜索用时 31 毫秒
1.
The use of digital camcorders in biomechanical analyses can introduce errors due to inter-camera time offsets. The faster the motion being recorded the greater the error. A sequential synchronization method was developed in this study to achieve sub-field camera synchronization for multiple camcorders through numerical optimization. A recreational golfer performed ten drives while being recorded with four digital camcorders (60-Hz sampling rate). Video signals were sampled at 10,000 Hz to determine the actual inter-camera time offsets. The optimized inter-camera time offsets were computed based on three markers placed on the shaft, each separated by 200 mm. The inter-camera time offset error was computed as the difference between the optimized and actual inter-camera time offsets. The inter-camera time offset error reduced on average from 0.518 to 0.019 fields (1 field = 16.7 ms) or less due to sequential optimization. The optimized global reconstruction errors were less than 19% of the unadjusted values. It was concluded that the ability to synchronize multiple (two or more) cameras using a sequential sub-field optimization strategy promises to extend the use of relatively inexpensive digital camcorders to motions considered too fast for the low field rates of such cameras. The sequential approach presented provides a balance between computation time and reconstruction accuracy.  相似文献   

2.
We aimed to compare multilayer perceptron (MLP) neural networks, radial basis function neural networks (RBF) and linear models (LM) accuracy to predict the centre of mass (CM) horizontal speed at low-moderate, heavy and severe swimming intensities using physiological and biomechanical dataset. Ten trained male swimmers completed a 7 × 200 m front crawl protocol (0.05 m.s?1 increments and 30 s intervals) to assess expiratory gases and blood lactate concentrations. Two surface and four underwater cameras recorded independent images subsequently processed focusing a three-dimensional reconstruction of two upper limb cycles at 25 and 175 m laps. Eight physiological and 13 biomechanical variables were inputted to predict CM horizontal speed. MLP, RBF and LM were implemented with the Levenberg-Marquardt algorithm (feed forward with a six-neuron hidden layer), orthogonal least squares algorithm and decomposition of matrices. MLP revealed higher prediction error than LM at low-moderate intensity (2.43 ± 1.44 and 1.67 ± 0.60%), MLP and RBF depicted lower mean absolute percentage errors than LM at heavy intensity (2.45 ± 1.61, 1.82 ± 0.92 and 3.72 ± 1.67%) and RBF neural networks registered lower errors than MLP and LM at severe intensity (2.78 ± 0.96, 3.89 ± 1.78 and 4.47 ± 2.36%). Artificial neural networks are suitable for speed model-fit at heavy and severe swimming intensities when considering physiological and biomechanical background.  相似文献   

3.
ABSTRACT

Professional American football games are recorded in digital video with multiple cameras, often at high resolution and high frame rates. The purpose of this study was to evaluate the accuracy of a videogrammetry technique to calculate translational and rotational helmet velocity before, during and after a helmet impact. In total, 10 football impacts were staged in a National Football League (NFL) stadium by propelling helmeted 50th percentile male crash test dummies into each other or the ground at speeds and orientations representative of concussive impacts for NFL players. The tests were recorded by experienced sports film crews to obtain video coverage and quality typically available for NFL games. A videogrammetry procedure was used to track the position and rotation of the helmet throughout the relevant time interval of the head impact. Compared with rigidly mounted retroreflective marker three dimensional (3-D) motion tracking that was concurrently collected in the experiments, videogrammetry accurately calculated changes in translational and rotational velocity of the helmet using high frame rate (two cameras at 240 Hz) video (7% and 15% error, respectively). Low frame rate (2 cameras at 60 Hz) video was adequate for calculating pre-impact translational velocity but not for calculating the translational or rotational velocity change of the helmet during impact.  相似文献   

4.
Sedentary behaviour (SB) has emerged as a modifiable risk factor, but little is known about measurement errors of SB. The purpose of this study was to determine the validity of 24-h Physical Activity Recall (24PAR) relative to SenseWear Armband (SWA) for assessing SB. Each participant (n = 1485) undertook a series of data collection procedures on two randomly selected days: wearing a SWA for full 24-h, and then completing the telephone-administered 24PAR the following day to recall the past 24-h activities. Estimates of total sedentary time (TST) were computed without the inclusion of reported or recorded sleep time. Equivalence testing was used to compare estimates of TST. Analyses from equivalence testing showed no significant equivalence of 24PAR for TST (90% CI: 443.0 and 457.6 min · day–1) relative to SWA (equivalence zone: 580.7 and 709.8 min · day–1). Bland–Altman plots indicated individuals that were extremely or minimally sedentary provided relatively comparable sedentary time between 24PAR and SWA. Overweight/obese and/or older individuals were more likely to under-estimate sedentary time than normal weight and/or younger individuals. Measurement errors of 24PAR varied by the level of sedentary time and demographic indicators. This evidence informs future work to develop measurement error models to correct for errors of self-reports.  相似文献   

5.
Determining an athlete’s speed from broadcast video is a common practice in sport. Many software packages that perform data extraction from video files are expensive; however, open source software is also available, but lacks published validation for speed measurements. The purpose of this research was to examine the error of speed measurements extracted from video during an ice hockey skating exercise using open source software. The subject completed four exercises, at two speeds recorded by broadcast cameras set at five angles. The speeds from the broadcast cameras were compared to speeds calculated from a high-speed camera placed orthogonally to the exercise. Speeds from the broadcast cameras correlated well with the high-speed video for motion more than 12 m away from the broadcast camera. When comparing all the measured speeds, no significant difference was found between the speeds calculated by the high-speed camera (slow: 4.46 m/s ± 0.2; fast: 7.2 m/s ± 0.7) and the speed calculated from the broadcast cameras (slow: 4.50 m/s ± 0.4; fast: 7.34 m/s ± 0.6) (p > 0.05). The open source method was found to be less accurate when the athlete was close to (within 12 m of camera position) or moving directly toward the broadcast cameras.  相似文献   

6.
In the current study, we quantified biological movement variability on the start and early acceleration phase of sprinting. Ten male athletes aged 17–23 years (100-m personal best: 10.87 ± 0.36 s) performed four 10-m sprints. Two 250-Hz cameras recorded the sagittal plane action to obtain the two-dimensional kinematics of the block start and initial strides from subsequent manually digitized APAS motion analysis. Infra-red timing lights (80 Hz) were used to measure the 10-m sprinting times. The coefficient of variation (CV%) calculation was adjusted to separate biological movement variability (BCV%) from estimates of variability induced by technological error (SEM%) for each individual sprinter and measure. Pearson's product–moment correlation and linear regression analysis were used to establish relationships between measures of BCV% and 10-m sprint start performance (best 10-m time) or 10-m sprint start performance consistency (10-m time BCV%) using SPSS version 12.0. Measurement error markedly inflated traditional measures of movement variability (CV%) by up to 72%. Variability in task outcome measures was considerably lower than that observed in joint rotation velocities. Consistent generation of high horizontal velocity out of the blocks led to more stable and faster starting strides.  相似文献   

7.
水下三维物空间坐标的重构精度分析   总被引：1，自引：0，他引：1  
目的:水下运动学数据的精度对于游泳项目的技术动作分析十分重要,利用防水摄像装置对标定框架上的有效点在水下和空气中进行三维重构,旨在验证不同测试环境和不同大小的标定空间对水下三维重构精度的影响,为水下游泳的器材研发提供一定的参考和依据。方法:2台水下摄像机同步拍摄放置在泳池里的标定框架(2 m×2 m×1 m),标定框架采取2种不同大小的标定空间,分别对同样数目的控制点(12个)和有效点(4个)分别在水下和空气中进行三维重构,重构精度通过4个有效点的均方根(RMS)重构误差来表示,利用DLT算法和ARIEL软件进行标定重构和数字化。结果发现,在整体标定空间控制点的RMS重构误差在X、Y、Z轴上为:(ⅰ)5.60 mm、10.44 mm和3.74 mm(水下),(ⅱ)4.89 mm、4.89 mm和3.96 mm(空气);局部标定空间为(ⅰ)5.29 mm、9.18 mm和3.02 mm(水下)(ⅱ)1.44 mm、1.11 mm和3.25 mm(空气)。结论:由于光线折射不论标定空间的大小,水下三维重构所得到的RMS重构误差都高于空气中的测量,Y轴最为明显;随着标定空间的增大,两种测试环境下的RMS重构误差均随之增大。  相似文献   

8.
Abstract

The objective of this study was to analyse the correctness of the offside judgements of the assistant referees during the final round of the FIFA 2002 World Cup. We also contrasted two hypotheses to explain the errors in judging offside. The optical error hypothesis is based on an incorrect viewing angle, while the flash-lag hypothesis refers to perceptual errors associated with the flash-lag effect (i.e. a moving object is perceived as spatially leading its real position at a discrete instant signalled by a briefly flashed stimulus). Across all 64 matches, 337 offsides were analysed using digital video technology. The error percentage was 26.2%. During the first 15 min match period, there were significantly more errors (38.5%) than during any other 15 min interval. As predicted by the flash-lag effect, we observed many more flag errors (86.6%) than non-flag errors (13.4%). Unlike the predictions of the optical error hypothesis, there was no significant difference between the correct and incorrect decisions in terms of the positioning of the assistant referees relative to the offside line (0.81 and 0.77 m ahead, respectively). To reduce the typical errors in judging offside, alternative ways need to be considered to teach assistant referees to better deal with flash-lag effects.  相似文献   

9.
ABSTRACT

Males (n = 29) and females (n = 34) performed vertical jumps. Jump height was estimated from force platform data using five numerical methods and compared using intraclass correlation (ρ), and linear and rank regression standard error of estimate (SEE). Take-off velocity plus center of mass height at take-off and mechanical work methods were identical (ρ = 1.000) with small linear (SEE < 0.0001 cm) and rank order (SEE = 0) error. Integrating acceleration to displacement was nearly identical (ρ > 0.999), with small linear (SEE = 0.1 cm) and rank order (SEE < 1) errors compared to the above methods. Take-off velocity (ρ = 0.517) and flight time (ρ = 0.573) estimates were less than all other methods and had large linear (SEE ≥ 2.1 cm) and rank order (SEE > 4) errors. Take-off velocity and flight time methods should be avoided due to systematic and random error.  相似文献   

10.
Abstract

This study seeks to discover whether handball goalkeepers employ a general anticipatory strategy when facing long distance throws and the effect of uncertainty on these strategies. Seven goalkeepers and four throwers took part. We used a force platform to analyse the goalkeeper's movements on the basis of reaction forces and two video cameras synchronised at 500 Hz to film the throw using 3D video techniques. The goalkeepers initiated their movement towards the side of the throw 193 ± 67 ms before the release of the ball and when the uncertainty was reduced the time increased to 349 ± 71 ms. The kinematics analysis of their centre of mass indicated that there was an anticipatory strategy of movement with certain modifications when there was greater uncertainty. All the average scores referring to velocity and lateral movement of the goalkeeper's centre of mass are significantly greater than those recorded for the experimental situation with bigger uncertainty. The methodology used has enabled us to tackle the study of anticipation from an analysis of the movement used by goalkeepers to save the ball.  相似文献   

11.
The aim of this study was to analyse the important kinematic variables in elite men's and women's 20 km race walking. Thirty men and 30 women were analysed from video data recorded during the World Race Walking Cup. Video data were also recorded at four points during the European Cup Race Walking and 12 men and 12 women analysed from these data. Two camcorders operating at 50 Hz recorded at each race for 3D analysis. The two main performance determinants of speed were step length and cadence. Men were faster than women because of their greater step lengths but there was no difference in cadence. A reduction in step length was the initial cause of slowing down with later decreases in speed caused by reductions in cadence. Shorter contact times were important in optimising both step length and cadence, and faster athletes tended to have longer flight times than slower athletes. It was less clear which other kinematic variables were critical for successful walking, particularly with regard to joint angles. Different associations were found for some key variables in men and women, suggesting that their techniques may differ due to differences in height and mass.  相似文献   

12.
FINA recently approved the backstroke ledge (Omega OBL2) to improve backstroke start performance in competition, but its performance has not been thoroughly evaluated. The purpose of this study was to compare the mechanics of starts performed with and without the OBL2. Ten high-level backstroke swimmers performed three starts with, and three starts without, the OBL2. A wall-mounted force plate measured the lower limb horizontal impulse, vertical impulse, take-off velocity and take-off angle. Entry distance, time to 10 m and start of hip and knee extension were recorded using video cameras. Starts performed with the OBL2 had a 0.13 s lower time to 10 m, 2.5% less variability in time to 10 m and 0.14 m greater head entry distance. The OBL2 provides a performance advantage by allowing an increased head entry distance rather than larger horizontal impulse on the wall. This may be due to the swimmers assuming different body positions during the start manoeuvre. Additional studies are needed to evaluate factors that contribute to improved performance when using the OBL2. Swimmers should train with the OBL2 and use it in competition to ensure optimal start performance.  相似文献   

13.
Mechanography during the vertical jump may enhance screening and determining mechanistic causes underlying physical performance changes. Utility of jump mechanography for evaluation is limited by scant test-retest reliability data on force-time variables. This study examined the test-retest reliability of eight jump execution variables assessed from mechanography. Thirty-two women (mean±SD: age 20.8 ± 1.3 yr) and 16 men (age 22.1 ± 1.9 yr) attended a familiarization session and two testing sessions, all one week apart. Participants performed two variations of the squat jump with squat depth self-selected and controlled using a goniometer to 80º knee flexion. Test-retest reliability was quantified as the systematic error (using effect size between jumps), random error (using coefficients of variation), and test-retest correlations (using intra-class correlation coefficients). Overall, jump execution variables demonstrated acceptable reliability, evidenced by small systematic errors (mean±95%CI: 0.2 ± 0.07), moderate random errors (mean±95%CI: 17.8 ± 3.7%), and very strong test-retest correlations (range: 0.73–0.97). Differences in random errors between controlled and self-selected protocols were negligible (mean±95%CI: 1.3 ± 2.3%). Jump execution variables demonstrated acceptable reliability, with no meaningful differences between the controlled and self-selected jump protocols. To simplify testing, a self-selected jump protocol can be used to assess force-time variables with negligible impact on measurement error.  相似文献   

14.
ABSTRACT

The predictive value of the multiple hop test for first-time noncontact lateral ankle sprains. Background: Lateral ankle sprains (LAS) are very common sports injuries, cause high health care costs and are associated with postural control deficits. From a preventive point of view, clinicians should dispose valid field tests to identify athletes at risk for a LAS. The aim of this study is to evaluate the predictive value of the multiple hop test (MHT) for first-time noncontact LAS. Methods: Non-elite athletes (n = 232) performed the MHT at baseline. During a 12-month follow-up period, all noncontact LAS related to health care costs were recorded. Outcomes of the MHT (completion time, balance errors and perceived difficulty) between the injured and uninjured group were compared and odds ratios (OR) and relative risks (RR) were calculated using a logistic regression analysis. Results: Ten first-time noncontact LAS were recorded (4.3%). Injured athletes made significantly more change-in-support strategy (CSS) errors when compared to uninjured athletes (p = .04). The OR of the number of CSS errors was 1.14 (p = .03), the RR 4.1 (p = .04). Conclusions: Athletes scoring > 12 CSS errors, have a four times increased risk for a first-time noncontact LAS. The MHT is a valid field test to identify athletes at risk for a first-time noncontact LAS.  相似文献   

15.
The objectives of the study were to describe lower limb kinematics in three dimensions during the forward skating stride in hockey players and to contrast skating techniques between low- and high-calibre skaters. Participant motions were recorded with four synchronized digital video cameras while wearing reflective marker triads on the thighs, shanks, and skates. Participants skated on a specialized treadmill with a polyethylene slat bed at a self-selected speed for 1 min. Each participant completed three 1-min skating trials separated by 5 min of rest. Joint and limb segment angles were calculated within the local (anatomical) and global reference planes. Similar gross movement patterns and stride rates were observed; however, high-calibre participants showed a greater range and rate of joint motion in both the sagittal and frontal planes, contributing to greater stride length for high-calibre players. Furthermore, consequent postural differences led to greater lateral excursion during the power stroke in high-calibre skaters. In conclusion, specific kinematic differences in both joint and limb segment angle movement patterns were observed between low- and high-calibre skaters.  相似文献   

16.
The purpose of this study was to determine the installation of the backstroke start device reduces 15-m time. Thirteen college swimmers participated in this study. The aerial start and underwater motions were recorded with two digital video cameras. The center of mass (CM) of the swimmer, angular displacements and velocities of the shoulder, hip and knee joints were calculated. As an indicator of performance, the 5- and 15-m times were measured. The 5- and 15-m times in the backstroke start device condition were significantly shorter than in the non-backstroke start device condition. The vertical velocities of the CM at hand-off and toe-off in the backstroke start device condition were significantly greater than in the non-backstroke start device condition, while there was no significant difference in the CM horizontal velocity at toe-off. As a result, the height of the great trochanter at entry of the fingertips, with the backstroke start device, was 15 cm higher than in the non-backstroke start device condition. In addition, the CM horizontal velocities at 5 m in the backstroke start device condition were significantly greater than those of the non-backstroke start device. Thus, the use of the backstroke start device may reduce the 15-m time by diminution of the entry area.  相似文献   

17.
We investigated the rotational effect of buoyant force around the body’s transverse axis, termed buoyant torque, during a 200m front crawl maximal swim. Eleven male swimmers of national or international level participated. One stroke cycle (SC) for each 50m was recorded with two above and four below water cameras. The following variables were analysed: swimming velocity; absolute and normalised buoyant force; minimum, average and maximum buoyant torque; SC and arm recovery times. The average value of buoyant torque was higher in the first 50m (14.2 ± 4.5Nm) than in the following 150m (9.3 ± 4.1Nm~10.9 ± 4.5Nm) and was directed to raise the legs and lower the head throughout the race. The change in its magnitude seemed to be linked to the shorter time spent proportionally in arm recovery (first 50m: 27.6% of SC time; next 150m: 23.3–24.4% of SC time). Most swimmers had periods of the SC where buoyant torque was directed to sink the legs, which accounted to 10% of SC time in the first 50m and about twice this duration in the next 150m. These periods were observed exclusively at some instances when the recovering arm had entered the water while the opposite arm was still underwater.  相似文献   

18.
Where as video cameras are a reliable and established technology for the measurement of kinematic parameters, accelerometers are increasingly being employed for this type of measurement due to their ease of use, performance, and comparatively low cost. However, the majority of accelerometer-based studies involve a single channel due to the difficulty associated with synchronizing multiple accelerometer channels. The authors of this article outline a method to synchronize multiple accelerometers using a maxima detection method. Results are presented that demonstrate the effectiveness of the new synchronization method with 52 of 54 recorded data sets showing no time lag error and two tests showing an error of .04 sec. Inter-instrument and instrument-video correlations are all greater than r = .94 (p < .01), with inter-instrument precision (Root Mean Square Error; RMSE) ≈ .1ms?2, demonstrating the efficacy of the technique. In conclusion, the new technique offers a robust solution, giving further support to the movement toward wider adoption of accelerometer-based performance measurement systems in sports science.  相似文献   

19.
The landing error scoring system (LESS) assesses the quality of a landing after a jump. The quality of the jump is usually evaluated using a three-dimensional (3-D) motion analysis system or a two-dimensional (2-D) video analysis visually rated by a clinician. However, both methods have disadvantages. The aim of this study was to examine the concurrent validity of a novel portable motion analysis system (‘PhysiMax System’) in assessing the LESS score by comparing it to video analysis. The study population included 48 healthy participants (28.45 ± 5.61 years), each performing the LESS test while two video cameras and the ‘PhysiMax’ simultaneously recorded the jump. The ‘Physimax’ system automatically evaluated the LESS. Subsequently, the examiners scored the test by viewing the video recordings, blinded to the ‘PhysiMax’ results. The mean LESS score, using the video recordings and the ‘PhysiMax’ was 4.77 (±2.29) and 5.15 (±2.58), respectively, (ICC = 0.80, 95% confidence intervals 0.65–0.87), mean absolute differences 1.13 (95% confidence intervals; 0.79–1.46). The results indicate a high consensus between the methods of measurement. The ‘Physimax’ system’s main advantages are portability, objective evaluation and immediate availability of results. The system can be used by athletic trainers and physiotherapists in the clinic and in the field for jumping assessment.  相似文献   

20.
The purpose of this study was to analyse the raw lifting speed collected during four different resistance training exercises to assess the optimal sampling frequency. Eight physically active participants performed sets of Squat Jumps, Countermovement Jumps, Squats and Bench Presses at a maximal lifting speed. A linear encoder was used to measure the instantaneous speed at a 200 Hz sampling rate. Subsequently, the power spectrum of the signal was computed by evaluating its Discrete Fourier Transform. The sampling frequency needed to reconstruct the signals with an error of less than 0.1% was f99.9 = 11.615 ± 2.680 Hz for the exercise exhibiting the largest bandwidth, with the absolute highest individual value being 17.467 Hz. There was no difference between sets in any of the exercises. Using the closest integer sampling frequency value (25 Hz) yielded a reconstruction of the signal up to 99.975 ± 0.025% of its total in the worst case. In conclusion, a sampling rate of 25 Hz or above is more than adequate to record raw speed data and compute power during resistance training exercises, even under the most extreme circumstances during explosive exercises. Higher sampling frequencies provide no increase in the recording precision and may instead have adverse effects on the overall data quality.  相似文献   

设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号