Autonomous sports training from visual cues (original) (raw)
Related papers
Visual tracking for sports applications
2005
Visual tracking of the human body has attracted increasing attention due to the potential to perform high volume low cost analyses of motions in a wide range of applications, including sports training, rehabilitation and security. In this paper we present the development of a visual tracking module for a system aimed to be used as an autonomous instructional aid for amateur golfers. Postural information is captured visually and fused with information from a golf swing analyser mat and both visual and audio feedback given based on the golfers mistakes. Results from the visual tracking module are presented.
A Proposal for Detection and Estimation of Golf Putting
This study presents an experimental research design of a PhD work, studying the effects of the variability in the performance of the Golf putting. The instruments used to analyze the putting were two digital cameras to detect the relevant dynamic objects (i.e., ball and putter) and a biaxial accelerometer to confirm the exact moment at which the putter hits the ball. To synchronize the instruments, it was used a trigger. The ball's trajectory and the putting movement were automatically analyzed based on visual detection and parameter estimation. The kinematic analysis G. Dias ( ) Faculty of Sport Sciences and Physical Education, of the putting was performed using the Matlab software, to determine the amplitude, velocity and acceleration of the players' gestures. We concluded that the Golf putting action parameters are divided into different stages (i.e., backswing, downswing and follow-through) and that those can be useful to investigate the effects of variability in this movement.
Development of Pattern Recognition Methods for Golf Swing Motion Analysis
informatik.uni-erlangen.de
The golf swing is one of the most complex movement sequences in any sports. Even human experts can be overwhelmed by the amount of details that have to be taken into account for its analysis. We present a novel pattern recognition approach that can help in this analysis by automatically and robustly evaluating even tiny swing differences. Our approach is based on the accurate 3D spatiotemporal information about the posture of the golfer and the position of the club that the TaylorMade MAT-T TM motion-capture based swing measurement system provides. Golf club fitting experts have been using these systems throughout the last decade and have captured more than 500 000 swings worldwide. Based on the positional data contained in this unique database, we developed a feature description of the golf swing with the goal of classifying even small differences between groups of players. In this manuscript, we show the results of the application of several classifiers to two selected problems of group classification. The presented system can be used for example to distinguish expert from novice players. The information that is calculated by our software tool can substantially support the process of golf club fitting and furthermore assist coaches and golfers to improve performance.
7th Vienna International Conference on Mathematical Modelling, 2012
Sensor and computing technologies provide people with information on their performance and load when doing sports. In order to automatically give advices on how to continue exercising and/or to adjust the sports equipment during the physical activity, intelligent devices are required. These devices rely on models for recognition and classification of patterns in the motion currently performed. Different methods and models, such as Neural Networks, Hidden Markov models or Support Vector Machines have proven to be applicable for this purpose. Pros and cons of the different approaches are discussed. Practical applications are presented and experiences reported.
A photogrammetric application in virtual sport training
Photogrammetric Record, 2009
The paper discusses an application of close range photogrammetry for the development of a virtual training system for rugby football, and the use of the technique for the evaluation of rugby players' performance. NuView, a stereoimaging device, and a digital high-definition video (HDV) camera were used to capture stereoscopic video footage of players during field training. The left view and the right view were colour-tinted cyan and red, respectively. The tinted stereo (anaglyph) views were projected onto a white screen, and players were instructed to practise ball-throwing at the screen. A custom-built laser device (TAM) measured the accuracy of the virtual throws. In addition, a photogrammetric system was used to track the movement of body segments, for example, the angle of shoulder orientation and the trunk flexion of the thrower. The measurements determined the parameters needed for an accurate throw and these parameters would be used in the training of new players. The study shows statistically significant differences in the values of these parameters between experienced and inexperienced players.
Visual Golf Club Tracking for Enhanced Swing Analysis
Procedings of the British Machine Vision Conference 2003, 2003
This paper presents a new visual tracking technology that relies on the use of a global motion model to achieve robustness. We demonstrate its effectiveness for the purpose of retrieving the 2D spatio-temporal trajectory of a golf club head from ordinary video sequences of golf swings, so that information about club orientation, local speed and acceleration can also be obtained. We have integrated it into a fully automated system that requires neither user intervention nor the use of instrumented golf club or clothing, and that is usable in a natural environment with a potentially cluttered background. Our algorithm robustly fits a global swing trajectory model to club location hypotheses obtained from single frames. This process makes our approach very robust and it will soon be integrated into a commercial product. Several experimental results are presented to illustrate the success of this new method.
A Pilot Study on Human Pose Estimation for Sports Analysis
Lecture notes in electrical engineering, 2022
Human pose estimation is the identification and detection of different poses of a human through the information collected from body part movements where the body parts refer to the joints and the bones. By referencing a video, it can calculate accurate poses and body movements for athletes so that they can accomplish optimum results in their performance. Pose Estimation can also be further used to identify the health condition of a particular player. We have developed a model which identifies various anatomical key points of a person in a given image or a video (frames) for Pose Estimation. We further attempt to extract insights on the body movement of an athlete to carry out analysis of their running behavior. The model accurately extracts 18 anatomical key points (like Hip Joint, Knee joint, Ankle Joint, etc.) without the need of any laboratory settings and special sensors, which makes it easy for anyone to use and implement the model. The model used is based on the MobileNet CNN architecture. For analysis we use various gait parameters such as Cadence, Knee angle, and Velocity. We further attempt to compare the results of the striding patterns with the running patterns shown by an athlete. The model is able to track the body movements of an athlete and then output the various gait parameters associated with these body movements. The implementation of the model has been made easy to assist the athletes in achieving optimal performance without the need of personal trainers and equipment, which can be quite costly.
ArXiv, 2022
Recent developments in video analysis of sports and computer vision techniques have achieved significant improvements to enable a variety of critical operations. To provide enhanced information, such as detailed complex analysis in sports like soccer, basketball, cricket, badminton, etc., studies have focused mainly on computer vision techniques employed to carry out different tasks. This paper presents a comprehensive review of sports video analysis for various applications: high-level analysis such as detection and classification of players, tracking player or ball in sports and predicting the trajectories of player or ball, recognizing the team‟s strategies, classifying various events in sports. The paper further discusses published works in a variety of application-specific tasks related to sports and the present researcher‟s views regarding them. Since there is a wide research scope in sports for deploying computer vision techniques in various sports, some of the publicly avail...
Contour tracking of human exercises
2009 IEEE Workshop on Computational Intelligence for Visual Intelligence, 2009
We developed a novel markerless motion capture system and explored its use in documenting elder exercise routines in a health club. This system uses image contour tracking and swarm intelligence methods to track the location of the spine and shoulders during three exercises-treadmill, exercise bike, and overhead lateral pull-down. Preliminary results of our qualitative study demonstrate that our system is capable of providing important feedback about the posture and stability of elders while they are performing exercises. Study participants indicated that feedback from this system would add value to their exercise routines.
Semi-automatic tracking of beach volleyball players
2012
With the increase of computational power in the last decades, tracking of objects or human beings has become a growing research area in the image processing and engineering fields. Its applications range from military and security to medicine and sports. In particular, tracking applications in sports have the main purpose of extracting useful information for the analysis of a player's or team's performance through video analysis based on statistical measures (Kristan et al. 2009).