Volleyball Activity Detection Using Filtered Acceleration Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current algorithms in portable electronic devices struggle to accurately differentiate between activities such as walking and running, and are even more challenging with complex activities like volleyball, where various techniques and impact forces need to be distinguished.
Innovation Solution
A method using a combination of high-pass and low-pass filters to detect significant activities, followed by data preparation and classification using a 1-Nearest Neighbor algorithm with L1-norm, and optionally determining smash power through band-pass filtering, to characterize volleyball techniques like serves, passes, and digs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If simple algorithms are used for activity detection, then device complexity is reduced, but measurement precision deteriorates
Solution Approach 1:
The algorithm divides activity detection into multiple stages: first detecting potential activities using high-pass filtering, then filtering maxima above thresholds, preparing data by low-pass filtering from both ends towards the hit, and finally classifying using 1-Nearest Neighbor algorithm. This segmentation allows each stage to focus on specific aspects, improving overall precision without requiring a single complex algorithm.
Solution Approach 2:
The method performs preliminary filtering actions before classification. High-pass filter detects fast changes, low-pass filter removes noise, and threshold filtering identifies potential hits before the actual classification step. This preliminary action reduces the computational burden on the classification algorithm while improving accuracy.
2Measurement precision
If complex algorithms are used to detect complex activities like volleyball, then measurement precision improves, but device complexity increases
Solution Approach 1:
The volleyball detection algorithm is segmented into distinct functional blocks: acceleration data acquisition, high-pass filtering for fast changes, low-pass filtering for noise reduction, threshold filtering for hit detection, data preparation with bidirectional low-pass filtering, and classification using 1-Nearest Neighbor. This modular approach makes the complex algorithm more manageable and implementable.
Solution Approach 2:
The patent introduces intermediary filtering steps between raw sensor data and final classification. The high-pass filter acts as an intermediary to detect fast changes, the low-pass filter serves as an intermediary to remove noise, and the threshold filtering acts as an intermediary to identify potential hits before classification. These intermediaries simplify the overall process.
3Measurement precision
If continuous data acquisition is performed, then measurement precision improves, but use of energy increases
Solution Approach 1:
The algorithm uses periodic threshold checking during data acquisition. Instead of continuously processing all data, the system acquires data continuously but only performs complex processing when acceleration exceeds a threshold, indicating a potential hit. This periodic activation of processing reduces energy consumption while maintaining detection accuracy.
Solution Approach 2:
The system performs preliminary threshold filtering before full data processing. By first checking if acceleration exceeds a threshold, the system can skip extensive processing for normal movements and only activate full analysis when a hit is detected, reducing overall energy consumption while maintaining precision.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
The method effectively differentiates and characterizes volleyball techniques, reducing non-relevant activity detection and providing accurate data for displaying technique occurrences and impact forces, enhancing the efficiency of portable electronic devices in tracking sports activities.
Implementation Method 1
an accelerator sensor (30) attached to a wrist
Implementation Method 2
Detecting potential activities using a high-pass filter the acquired data so as to detect fast changes of the measured acceleration
Implementation Method 3
using a low pass filter to filter the norm of the high-passed filtered data to only detect significant activities
Implementation Method 4
it further includes a step of determining the hit power consisting in using a band-pass filter to filter acceleration data in a certain window before and after the hit
Data Source
AI summary
The present invention concerns a method of detection for an electronic apparatus comprising a case containing an electronic module powered with electrical energy by a means of storing electrical energy, said electronic module including a calculation unit connected to an accelerator sensor and to a memory unit, said electronic module also being connected to a display means so as to display information linked to some data of said accelerator sensor.