Image Edge Detection Methods in Perimeter Security Systems Using Distributed Fiber Optical Sensing.

Area of Science:

• Optoelectronics
• Sensor Technology
• Signal Processing

Background:

• Perimeter security systems require robust intrusion detection.
• Phase-sensitive optical time-domain reflectometry (Φ-OTDR) is a key technology for such applications.
• Existing detection algorithms may have limitations in performance and integration.

Purpose of the Study:

• To evaluate edge detection algorithms for perimeter intrusion detection using Φ-OTDR.
• To compare the performance of Sobel and Prewitt operators against conventional methods.
• To assess the effectiveness of a novel integrated Φ-OTDR sensor system.

Main Methods:

• Development of a compact 1U rack-mounted Φ-OTDR sensor system with FPGA-based signal processing.
• Testing the system on a polygon with varied fence types and underground cables.
• Application of Sobel and Prewitt edge detection algorithms for post-processing.
• Comparison of algorithm performance using signal-to-noise ratio (SNR) metrics.

Main Results:

• Edge detection algorithms, specifically Sobel and Prewitt, were successfully applied to Φ-OTDR data.
• The performance of these algorithms was quantified using SNR values for different intrusion events.
• A comparison indicated potential advantages of edge detection methods over the conventional differential method.

Conclusions:

• Edge detection algorithms offer a viable alternative for intrusion detection in Φ-OTDR perimeter security systems.
• The developed integrated Φ-OTDR system demonstrates effective performance in diverse scenarios.
• Further research can optimize these algorithms for enhanced perimeter security solutions.