基于像素比的图像型感烟火灾探测器灵敏度检测

Smoke-based video fire detectors sensitivity tests based on pixel aspect ratio

针对新版国家标准《图像型火灾探测器》(报批稿)中新增的灵敏度检测要求,研究典型ISO标准试验火的像素比特性和灵敏度检测方法。利用光学密度计、离子烟浓度计、烟雾视频测试装置等,分别对木材热解阴燃火(TF2)、棉绳阴燃火(TF3)和聚氨酯塑料火(TF4)进行数据采集与数据关联。通过高斯滤波、灰度分段线性变换和区域生长算法,对视频图像进行烟雾识别并计算像素比。通过高次线性最小二乘拟合,对多组试验火的光学浓度、离子浓度及像素比之间的对应关系及分段特性进行分析。对采用TF2、TF3和TF4标准试验火进行VFD/S灵敏度检测的试验终止条件提出了建议。

In the new national standards Video Fire Detectors （draft）, a new sensitivity test is introduced. Considering the new requirement, researches have been done on PAR character- istics and sensitivity test methods using typical ISO standard test fires. Using optical densitometers, ion densitometers and smoke video test equipment, data collection and data association was conducted in the smouldering pyrolysis wood fire （TF2）, smouldering cotton fire （TF3） and （）pen plastics fire （TF4）, respectively. Using Gaussian filtering, gray-scale subsection linear transformation and region growing algorithm, the smoke in the video image was identified and the PAR was calculated. Using high-order linear least squares fitting, the correspon- dence and distribution characteristics between the optical densi- ty, the ion density and PAR was analyzed. Recommendations on termination condition of VFD/S sensitivity tests using TF2, TF3 and TF4 have been proposed.