Personalised annotations for augmented video coaching
Modern mobile video and camera technologies can provide high frame rates and high-quality videos. Such video technologies can advance augmented coaching and support qualitative movement diagnostics to help sports enthusiasts to improve their golf swing. This thesis presents novel approaches and technology for privacy-preserving personalised annotations for augmented video coaching using golf as a case study. Preliminary experiments indicated that the commonly-used foreground-background separation algorithms would not perform well in golf-specific contexts. Initial evaluations included benchmarking and combining commonly used surveillance algorithms (using Matlab) that could provide a silhouette of a golfer and a club. Evaluated solutions combined frame difference, erosion/dilatation, blob detection and Gaussian mixture models from the captured video at two diverse-characteristic golf driving ranges. The adaptive multi-layered solution for privacy preservation of golfing activity can provide pseudo-3D binary silhouette transformation of video/image that can be used for augmented coaching and providing anonymised visual annotation feedback while preserving players’ privacy. Producing a video or generating a report with annotated angles, golf club head trajectory and other elements of swing performance are important coaching tools to facilitate golf learning from novice to intermediate skill level players. In addition, future work is aimed at further advancements of silhouette-based video streaming solutions and technology transfer to advance the diversity of sports disciplines, and general sports science.