【摘要】 为满足油气管道中甲烷气体泄漏的监测和定位需求，避免现有激光遥测式点测量分析的不足，实现非接触式红外气体泄漏的成像检测，提出一种基于双波段短波红外气体差分比值光谱成像的技术。在系统中，锁定甲烷测量时的激光输出中心波长，通过调整光束扫描模块实现激光束扫描轨迹与方向的实时控制，基于激光吸收光谱技术结合红外短波相机，采集有无吸收区域的光谱信息，进而生成不同体积分数甲烷气体在特定激光波长下的强度图像。利用图像处理算法，对甲烷气体吸收波段（1653.72 nm）和非吸收波段（1653.82 nm和1653.62 nm）的双波段进行差分比值分析，有效消除了甲烷气袋（团）处的背景辐射影响，清晰展现了不同体积分数甲烷气体的强度图。在实验中，通过标准气袋模拟甲烷泄漏分布情况，处理后的图像强度值用于生成双波段差分比值与气体浓度长度的拟合标定曲线，进而实现定量检测与分析。该成像方法有效弥补了传统测量技术测量点偏移、无法定量等不足，为精确定位矿井瓦斯云团分布、油气甲烷泄漏等提供了新的技术思路。更多还原

【Abstract】 To address the challenges associated with methane gas leak monitoring and localization in oil and gas pipelines, and to overcome the limitations of existing laser-based point measurement techniques, this study proposes a novel dualband short-wave infrared(SWIR) gas detection technology based on differential ratio spectral imaging. The system locks the laser output at a central wavelength specific to methane detection and controls the laser beam scanning trajectory and direction in real time through an adjustable beam scanning module. By integrating laser absorption spectroscopy with a shortwave infrared camera, spectral information from both absorption and non-absorption regions is captured to generate intensity images of methane gas at varying volume fractions under specific laser wavelengths. Through the application of image processing algorithms, differential ratio analysis is conducted between the methane absorption band(1653.72 nm) and two adjacent non-absorption bands(1653.82 nm and 1653.62 nm), effectively eliminating background radiation interference in methane gas plumes and distinctly visualizing the intensity distribution of methane volume fractions. Experimental validation using standard gas bags to simulate methane leak scenarios demonstrates the feasibility of this approach, with processed image intensity to establish a calibration curve correlating the dual-band differential ratio with gas concentration path length, enabling quantitative detection and analysis. This imaging methodology effectively mitigates the drawbacks of conventional measurement techniques, such as measurement point deviation and the inability to achieve quantitative assessment, thereby providing a robust and precise approach for applications including methane cloud distribution mapping in mining operations and methane leak detection in oil and gas fields.更多还原