This is kinda funny. Like, if I were facing this same problem, I’d have a completely-different way of going about doing that analysis – take footage from fixed camera during tennis matches, edit out the cuts to other cameras where people are doing things like walking around, and then just composite all the images in the remaining video. Maybe a subtract a (potentially-composite) image of the field without people on it to get just the people moving around. Would not have thought of looking at plain old real-world analog grass wear.
To me it seems less like they wanted to analyze the patterns of play and chose to look at grass over video, but instead just noticed the grass pattern change and ascribed to it a change in play patterns.
Hah!
This is kinda funny. Like, if I were facing this same problem, I’d have a completely-different way of going about doing that analysis – take footage from fixed camera during tennis matches, edit out the cuts to other cameras where people are doing things like walking around, and then just composite all the images in the remaining video. Maybe a subtract a (potentially-composite) image of the field without people on it to get just the people moving around. Would not have thought of looking at plain old real-world analog grass wear.
To me it seems less like they wanted to analyze the patterns of play and chose to look at grass over video, but instead just noticed the grass pattern change and ascribed to it a change in play patterns.