Development of a simple algorithm to detect big air jumps and jumps during skiing

Area of Science:

• Sports Science
• Biomechanics
• Sensor Technology

Background:

• Jumping is a critical performance metric in numerous sports, including skiing.
• Understanding jump mechanics is vital for athlete training and performance enhancement.
• Existing methods for jump analysis may lack simplicity or ease of implementation.

Purpose of the Study:

• To develop a straightforward jump detection algorithm for skiing.
• To utilize acceleration data from inertial measurement units (IMUs) in ski boots.
• To provide insights into different jump types in skiing.

Main Methods:

• Algorithm development based on average absolute vertical acceleration from dual ski boots.
• Testing the algorithm on Big Air jumps and two categories of skiing jumps (small and medium).
• Evaluation of algorithm performance across various parameter settings.

Main Results:

• The algorithm achieved 100% detection for Big Air jumps.
• Detection rates were 94% for medium jumps (time of flight ≥ 500 ms) and 44% for small jumps (time of flight < 500 ms).
• A trade-off was observed between detection rate and overdetection, leading to the proposal of a penalty-adjusted score.

Conclusions:

• The proposed algorithm offers a simple and effective method for detecting significant jumps in skiing.
• A penalty-adjusted score is recommended to balance detection accuracy with overdetection in performance analysis.
• This approach can aid coaches and athletes in analyzing jump performance more effectively.