App re-build with iOS 13.0 SDK to comply with latest App Store requirements.
“Tennis Serve Tracker” measures tennis serve speed using video processing and image recognition technologies. Why is it worth trying it when there are similar apps on market? Because all these apps can be split on 2 categories - first are the ones that do it very, very bad or simply invent measured speed and won’t be mentioned here at all. The others do measure the serve speed, but it’s a lot of manual interactions with setting multiple parameters, like choosing deuce/ad serve type, server height, marking server and ball positions on the standard court image template that has no resemblance to the actual court we see with our eyes. These actions drastically increase an error margin, time and effort required to perform a single serve speed measurement - does it match fast pacing tennis game we all love?
So how “Tennis Serve Tracker” measures tennis serve speed? The fundamental difference is that it processes every single frame in the video feed and recognizes the court on the image frame. It allows to reconstruct 3D world coordinates of the server and ball from the image coordinates. Application provides constant feedback by highlighting court lines, net and a ball on the screen. Now having the time measurements from the video and ball world coordinates application can calculate your serve speed with a great accuracy, most times way better than a commercial speed gun (see doppler effect cosine factor error)! Visit “Demo Videos” application website page for videos with real time side by side comparison between speed radar gun and the app.
Final speed display is unlocked by lifetime "Premium" upgrade just for a price of a single tennis balls can. This app is a valuable tool for every tennis player who would like to improve their game and have a great serve!
- Added short camera flash just before recognition starts to provide visual feedback to the server;
- Added recognition delay feature to start service tracking after predefined timeout in settings. Combined with "Burst" mode allows recognitions taken in the loop automatically without user interaction - single person can use the app with a tripod now.
- Re-implemented parts of the recognition algorithm to significantly increase successful automatic recognition rate;
- Added "demo" recognitions into the new app installations;
- Fixed "Help & FAQ" web links in iOS 9;
- Multiple bug fixes and application functionality improvements.