Multi-Modal Imaging System for Real-Time Liquid Leak Detection

用于实时液体泄漏检测的多模态成像系统

• 批准号: 543694-2019
• 负责人: Liu, Zheng
• 金额: $ 1.24万
• 依托单位: University of British Columbia
• 依托单位国家: 加拿大
• 项目类别: Collaborative Research and Development Grants
• 财政年份: 2021
• 资助国家: 加拿大
• 起止时间: 2021-01-01 至 2022-12-31
• 项目状态: 已结题
• 来源: https://www.nserc-crsng.gc.ca/ase-oro/Details-Detailles_eng.asp?id=719883
• 关键词: Multi; Modal; Imaging; System; Real

The oil and gas pipeline integrity management program (IMP) is to ensure the safe operation of the energy transportation infrastructure. The early detection of a small leak from above-ground facilities is a key element of the IMP and paramount for environmental safety and protection. The challenges come from the complicated situations of liquid leaks in a dynamic environment. An intelligent leak detection system, which is capable of detecting those leaks accurately and timely, is expected to address this issue. This research aims to develop computational algorithms, which can enhance the machine vision based leak detection with higher accuracy and lower false alarm rate. This intelligent vision system, which is comprised of visible and infrared imaging, will employ the advanced deep learning techniques and industry-level hardware to enable continuous and reliable monitoring of the pipeline infrastructure. Continuous monitoring allows detecting leaks quicker, reducing the environmental impacts from leaks and improve pipeline performance in high-risk areas. Thus, this research will contribute to the healthy growth of Canada's energy sector that provides countless economic benefits to Canadians across the country.

石油和天然气管道完整性管理计划（IMP）是为了确保能源运输基础设施的安全运行。及早发现地面设施的小泄漏是国际监测方案的一个关键要素，对环境安全和保护至关重要。挑战来自于动态环境中液体泄漏的复杂情况。一个智能泄漏检测系统，这是能够准确和及时地检测到这些泄漏，预计将解决这个问题。 本研究的目的是开发计算算法，可以提高基于机器视觉的泄漏检测的准确性和低误报率。该智能视觉系统由可见光和红外成像组成，将采用先进的深度学习技术和行业级硬件，以实现对管道基础设施的连续可靠监控。连续监测可以更快地检测泄漏，减少泄漏对环境的影响，并提高高风险地区的管道性能。因此，这项研究将有助于加拿大能源部门的健康发展，为全国各地的加拿大人提供无数的经济利益。