Advancing the measurement of motion in sport using markerless motion capture technology

体育科学家经常分析运动员完成特定运动任务时的运动模式，例如冰球运动员的滑冰。玩家的组进行比较（例如新手与精英），使科学家能够确定最佳的运动模式和组之间的差异。这些信息可以被教练用来设计训练计划，也可以被工业用来设计设备。体育科学家目前使用的技术是在球员身上放置许多小的反射球，称为标记。特殊的摄像机测量这些标记的位置，然后科学家们在运动员移动时构建3D图像。这项技术有其局限性，包括成本高，当标记没有正确放置在球员身上时会出现错误，并且收集和处理这些数据非常耗时。 一种不需要反射标记的新技术已经出现，称为无标记运动捕捉。它使用一种深度学习，计算机在人类移动的先前图像上进行训练。该软件使用这些信息和来自摄像机的图像来识别运动员在完成运动任务时的身体位置。这些下一代无标记系统可以产生准确的数据，同时具有减少数据收集和处理时间的额外好处，它们需要更少的专业知识来操作，并且成本更低。无标记运动捕捉已被用于测量步行，但其在体育运动中的使用将是新颖的。这项技术将支持三名研究人员的研究计划，包括研究1）足球运动员在跑步时应对障碍物的快速方向变化，2）男女冰球运动员的滑冰和射击，以及3）在不同的户外表面上跑步。其他研究人员和运动队将通过合作获得它，学生将有机会使用这项技术进行训练。无标记动作捕捉系统是我们现有设备的重大升级。由于无标记动作捕捉更有效，我们可以通过增加注册的玩家数量和允许我们完成更多研究来扩展我们的研究计划并提高生产力。当我们与运动队合作完成现场测试时，这种效率以及更高的便携性非常重要。这项技术将使我们能够培训更多的学生，因为它节省了时间，需要更少的专业知识。这些学生将成为这一前沿技术的专家，这将使他们在工业和学术界就业。无标记运动捕捉是一个新兴的领域，将改变体育科学。我们的研究将大大推进评估运动的研究。研究结果可用于教练设计训练计划，治疗师设计伤害预防计划，以及行业改进设备设计。这将有助于最大限度地提高加拿大运动员的表现。

Sport scientists frequently analyze the movement patterns of players when they complete sport specific tasks, such as skating in ice hockey players. Groups of players are compared (e.g. novice vs. elite), allowing scientists to determine optimal movement patterns and between group differences. This information can be used by coaches in designing training programs and by industry to design equipment. Sports scientists currently use a technology where numerous, small reflective spheres, called markers, are placed on the player. Special cameras measure the position of these markers and scientists then construct a 3D image of the players as they move. This technology has limitations, including it is costly, errors arise when the markers are not placed correctly on players, and it is time consuming to collect and process this data. A new technology has emerged which does not require reflective markers, called markerless motion capture. It uses a type of deep learning, where the computer is trained on previous images of humans moving. The software uses this information and images from video cameras to identify a player's body position when they are completing a sport task. These next-generation markerless systems can produce accurate data, while having the added benefit of decreasing data collection and processing times, they require less expertise to operate, and they have lower costs. Markerless motion capture has been used to measure walking but its use in sports will be novel. This technology will support the research programs of three researchers, including studies examining, 1) quick directional changes while running in response to obstacles in soccer players, 2) skating and shooting in male and female ice hockey players, and 3) running on different outdoor surfaces. Other researchers and sports teams will have access to it through collaborations, and students will have the opportunity to train with this technology. The markerless motion capture system is a significant upgrade from our current equipment. Since markerless motion capture is more efficient, we can expand our research programs and improve productivity by increasing the number of players enrolled and by allowing us to complete more studies. This efficiency, along with its increased portability, is important when we collaborate with sports teams to complete on-field testing. This technology we will allow us to train more students since it saves time and requires less expertise. These students will become experts with this leading edge technology, which will make them employable within industry and academia. Markerless motion capture is an emerging field that will transform sports science. Our studies will substantially advance research evaluating motion in sport. The findings can be used by coaches to design training programs, therapists to design injury prevention programs, and industry to improve equipment design. This will help maximize the performance of Canadian athletes.