Open 2gauravc opened 3 years ago
The code assumes, the video is in order squat -> peak -> landing.
In this one, T starts standing tall. Hence the below line causes squat frame to be the same as the peak frame.
min_squat_forehead_y_movement = np.nanmin(np.array(forehead_y_movement[:self.jump_peak_frame], dtype=np.float64)).astype(int)
Reference Video vjump-t1a.mov
LOG OUTPUT
03-Oct-20 09:45:04 - DEBUG - Squat posture frame = 1 03-Oct-20 09:45:04 - DEBUG - Jump Peak posture frame = 1 03-Oct-20 09:45:04 - DEBUG - Landing posture frame = 58
See attached for Frame 1. It is not close to Squat or Jump Peak position