Hey guys. As part of a project, I am looking for a feasible way of detecting the direction of a golf swing (backswing or downswing). The golfer could be either hitting a golf ball or doing a practice shot.
The measurement has to be done in a noninvasive manner. By noninvasive, I mean I can't shot a laser towards the golf club directed at a photodetector to see if the club is in the way or not. The golfer has to be able to set up this device as easily as possible, so no attachments can be made either to the clubs, balls, or tees.
My idea: use an array of two HC SR04 ultrasonic sensors facing the golfer (we use two sensors to be able to detect the swing's direction). The problem with these sensors is that they are only good at detecting objects that are large, flat and moving fairly slowly. Golf clubs are very thin and they travel at tens of mph when the golfer swings, so I haven't had much success using them. My only hope is to be able to detect at least the club's head, but that would make the device too dependant on the way the golfer swings, and the sensors would have to be dangerously close to the ball.
The NewPing library is excellent and is allowing me to get the most out of these ultrasonic sensors, but their physical limitations are making me think of other alternatives.
IR Sharp sensors are very fast and could detect thinner objects (maybe) but they don't work well outdoors, and in order to overcome this limitation the sensor would have to be too close to the golfer, making him uncomfortable and putting the sensor at risk.
I have been thinking of using LIDAR concepts since I don't need accuracy to measure distance, only an object's presence.
As a last resort, maybe I can use openCV, a webcam and a Raspberry PI, but that seems to be an overkill for what I want to do. ¿Any ideas? ¡Thank you people!