While many of us were watching the action at the Winter Olympics in Beijing, CAMERA Research Associate Laurie Needham and CAMERA PhD student Mike Muckelt (Uni of Bath/BBSA funded) headed off to attend a BBSA (British Bobsleigh and Skeleton Association) training camp in Sigulda, Latvia. The facility at Sigulda has a full ice track and an icehouse for practicing starts.

Previously the team have published work validating their high-speed camera systems on sprint tracks and the University’s dry land push track (see links below). The purpose of this trip was to test the athlete and sled tracking algorithms in ice environments and to collect new data that will allow them to update their systems to be more robust for data collected on ice. This in turn will allow the team to feed information back to BBSA coaches so they can start to understand on ice performance and also understand the differences between dry land pushing and on ice pushing. 

However collecting data in this environment was quite challenging! For example, calibrating cameras on ice, or having to put coats over the server to keep it warm enough in the -10 degrees weather are definitely a change from the usual conditions in Bath. 

Sensors | Free Full-Text | Can Markerless Pose Estimation Algorithms Estimate 3D Mass Centre Positions and Velocities during Linear Sprinting Activities? (mdpi.com)

CVPR 2020 Open Access Repository (thecvf.com)

“USING COMPUTER VISION AND DEEP LEARNING METHODS TO CAPTURE SKELETON PU” by Laurie Needham, Murray Evans et al. (nmu.edu)

Development, evaluation and application of a novel markerless motion analysis system to understand push-start technique in elite skeleton athletes (plos.org)