Abstract

Candidate Bib recognition in Physical Endurance Tests has been improved using AI-based image processing, RFID timing systems, and race timing technology. With the integration of Artificial Intelligence and UHF RFID tracking, data collection and performance tracking are now faster and more accurate. This system supports chip timing systems and event timing solutions, delivering real-time race tracking with high speed and reliability.

Problem

Detecting impersonation during endurance or recruitment events is challenging without automated race timing technology or RFID athlete tracking. Manual verification often disrupts athletic event management, leading to confusion in participant tracking and result accuracy. The absence of AI algorithms and chip timing systems makes maintaining reliable performance analysis difficult during large-scale tests.

Solution

We introduced an AI-based system integrated with RFID race timing solutions to recognize Bib numbers and verify participants in real time. The process combines computer vision and RFID timing systems to detect, match, and confirm whether the same registered candidate wears the assigned Bib number throughout the test. Supported by UHF RFID tracking, the system enhances event timing, race management, and performance tracking with dependable results.

Then the recognition process is carried out by determining whether the assigned Bib number is worn by that registered individual by referring to our previously collected candidate registration database.

Functions:

1. Dataset collection: Candidate videos from endurance events are captured and converted into images for training AI algorithms and RFID tracking systems.

2. Data annotation: Using LabelImg, Bib numbers are labeled for accurate participant tracking through AI and race timing technology.

The annotated objects are stored in an XML file and are reinspected with a Python script.

3. Data splitting: Annotated data is divided into training, validation, and testing to improve performance analysis.

4. Data training: The TensorFlow model trains the system to identify Bib numbers quickly, aligning with chip timing systems and UHF RFID tracking.

5. Monitoring the progress: Through TensorBoard, AI algorithms are refined for consistent event timing solutions and race management

6. Data testing: Final testing validates RFID athlete tracking and confirms reliable performance tracking results.

Tools and Software used:

LabelImg

Machine Learning

TensorFlow

AI Algorithms

RFID Race Timing Systems