开放空间传感器气瓶放气测试数据集

开放空间激光甲烷气云成像传感器针对传统激光云台形式单一点式测量容易产生误报、误测、测不到、看不见等不足，瞄准天然气泄漏定位难的痛点及定量分析的技术空白，项目从“点、线、面、体”四方面综合考虑，攻克激光光束扫描模组、光强自适应处理方法、风速羽流校正模型、质量流量计算方法等难题，率先研发了“开放空间天然气检测成像传感器”，实现“定量、定形、定位、定量”的测量。本数据集来自甲烷气体泄漏场景的模拟核心实验采集：通过高精度甲烷气瓶模拟实际工业或环境中的气体泄漏过程，精准控制排气速率与泄漏量以还原真实工况，采用传感器在不同距离时对泄漏区域进行动态检测与图像采集，并清晰生成成像检测结果，主要参数包括浓度梯度、扩散范围及聚集区域，为后续泄漏溯源与风险评估提供直接的数据支撑。本数据集包含两种文件：jpg 格式文件为实验过程中的静态成像图，该格式图片能够清晰定格特定时间节点的检测状态，为静态分析提供精准依据；mp4 格式文件则为动态视频，它能完整记录甲烷气云从泄漏、扩散到稳定的全过程，让甲烷气云的测量结果更生动、全面。上述数据的数据量约为68.51MB。Open-space laser methane gas cloud imaging sensor addresses the shortcomings of traditional single-point laser pan-tilt measurement, including false alarms, mismeasurements, missed detections and blind spots, targeting the pain points of difficult localization of natural gas leaks and the technical gap in quantitative analysis. The project comprehensively considers four dimensions: "point, line, plane and volume", overcoming technical challenges such as laser beam scanning modules, adaptive light intensity processing methods, wind-induced plume correction models and mass flow calculation methods, and takes the lead in developing the "open-space natural gas detection imaging sensor", realizing measurements with quantification, shape determination, position localization and quantification capabilities. This dataset is collected from core simulation experiments of methane gas leakage scenarios: High-precision methane gas cylinders were used to simulate the gas leakage process in actual industrial or environmental settings, with exhaust rate and leakage volume precisely controlled to restore real working conditions. The sensor was used to perform dynamic detection and image acquisition on the leakage area at different distances, and clearly generate imaging detection results. The main parameters include concentration gradient, diffusion range and aggregation area, providing direct data support for subsequent leakage source tracing and risk assessment. This dataset contains two types of files: JPG format files are static imaging images during the experiment. This format can clearly capture the detection status at specific time nodes, providing accurate basis for static analysis. MP4 format files are dynamic videos, which can fully record the entire process of methane gas cloud from leakage, diffusion to stabilization, making the measurement results of methane gas cloud more vivid and comprehensive. The total data volume of the above datasets is approximately 68.51 MB.

搜集汇总

背景与挑战

背景概述

该数据集聚焦于天然气泄漏检测，利用开放空间激光甲烷气云成像传感器，通过模拟甲烷气瓶泄漏实验采集数据。数据包含静态图像和动态视频，记录了不同距离下的气云浓度梯度、扩散范围等信息，为泄漏溯源与风险评估提供支持，总数据量约为68.51MB。