Visual Leakage Inspection in Chemical Process Plants Using Thermographic Videos and Motion Pattern Detection

Area of Science:

• Process Safety Engineering
• Non-Destructive Testing
• Infrared Imaging Technology

Background:

• Liquid leakage from pipelines presents significant operational and safety risks in chemical plants.
• Early detection of leaks is essential to prevent severe damage and hazardous situations.
• Vision-based inspection systems offer a promising approach for accurate and timely leakage detection.

Purpose of the Study:

• To develop and evaluate a vision-based system for early-stage liquid leakage detection in industrial pipelines.
• To utilize infrared (IR) imaging to capture thermal anomalies associated with leaking fluids.
• To employ motion pattern detection for identifying and differentiating leaking drops from background noise.

Main Methods:

• A Kalman filter-based approach was proposed to track the motion patterns of leaking drops.
• IR imaging was used to capture videos of simulated pipeline leakages in a laboratory setting.
• Learned motion patterns from training data were applied to test data for method validation.

Main Results:

• The proposed Kalman filter method successfully tracked the motion of leaking drops.
• The system demonstrated the ability to differentiate leaking drops from noise based on their motion patterns.
• The study confirmed the effectiveness of the IR imaging and motion tracking approach for leak detection.

Conclusions:

• The developed IR imaging and Kalman filter-based method provides an effective solution for early liquid leakage detection in pipelines.
• The approach shows potential for real-world application in industrial chemical plants, enhancing safety and operational efficiency.
• Further investigation into the conditions and implementation strategies for industrial deployment is recommended.