ReFlex is a program that utilizes machine learning capabilities, along with Mediapipe and OpenCV, to create a pose estimation model used to detect if a runner has false started in an Olympic-style track race.
In Olympic-style track races, such as sprints, false starts are a critical issue that can result in disqualification. ReFlex addresses this problem by leveraging machine learning techniques and the OpenCV library to create a pose estimation model. This model analyzes the movements of runners at the start line to determine if a false start has occurred.
Features:
-
Pose Estimation: Utilizes mediapipe and machine learning techniques to estimate the poses of runners at the start line.
-
False Start Detection: Detects if a runner has false started based on their initial movements, along with detected gunshot time.
-
Real-time Processing: Capable of processing video feeds in real-time for immediate feedback.
-
Customization: Easily adjustable parameters to fine-tune the detection algorithm for different race scenarios.