Deep localization and modeling of play-fields

运动场的深度定位和建模

• 批准号: 532178-2018
• 负责人: Yi, KwangMoo
• 金额: $ 1.71万
• 依托单位: University of Victoria
• 依托单位国家: 加拿大
• 项目类别: Engage Grants Program
• 财政年份: 2018
• 资助国家: 加拿大
• 起止时间: 2018-01-01 至 2019-12-31
• 项目状态: 已结题
• 来源: https://www.nserc-crsng.gc.ca/ase-oro/Details-Detailles_eng.asp?id=669329
• 关键词: Deep; localization; modeling; play; fields

Sport analytics has shown to be very effective for improving the performance of a team, or to even detect and**prevent injuries. SPORTLOGiQ is a market-leading Candian company that provides Computer Vision-based**sport analytic tools. In this research project, SPORTLOGiQ and the University of Victoria will collaborate for**the first time to broaden the scope of Artificial Intelligence (AI) assisted sport analytic tools.**Specifically, we will develop a Deep Neural Network that identifies the play-field, and use its results to**simultaneously track the camera pose as well as estimate the field dimensions. This will allow us to provide**accurate Augmented Reality (AR) for broadcast videos, and be able to analyze what is going on in the game,**for any sport out there.

体育分析已被证明对提高球队的表现非常有效，甚至可以检测和**防止受伤。SPORTLOGiQ是一家市场领先的加拿大公司，提供基于计算机视觉的**体育分析工具。在这个研究项目中，SPORTLOGiQ和维多利亚大学将首次合作**扩大人工智能(AI)辅助运动分析工具的范围。**具体地说，我们将开发一个识别比赛场地的深度神经网络，并使用其结果**同时跟踪摄像机姿势和估计场地尺寸。这将使我们能够为广播视频提供**准确的增强现实(AR)，并能够分析比赛中发生的事情，**针对任何运动。