基于视频的飞机货舱烟雾识别去干扰方法研究

Research on De-interference Method Based on Video for Aircraft Cabin Smoke Identification

针对传统飞机货舱烟雾探测器虚警率高和在低照度情况下检测结果易受环境干扰的问题,利用主动红外CCD摄像机进行视频监测,提出基于视频烟雾图像处理与去除灰尘干扰方法作为补充警报验证的方法。该方法实现货舱的自动烟雾识别:首先利用四帧差分法提取运动目标,采用中值滤波去除随机噪声,然后对图像连通域分割和质心标记提取,根据质心坐标拟合直线的斜率值,结合烟雾和灰尘运动方向的特点,利用SVM分类器将烟雾和灰尘区分以提高烟雾识别的准确度。在模拟飞机货舱内进行真实火源和灰尘干扰实验,实验结果表明,基于视频的飞机货舱烟雾识别方法可有效去除图像噪声和灰尘的影响,并进行烟雾早期预警。

Aiming at the high false alarm rate of traditional aircraft cargo tank smoke detector and the environmental interference caused by the low illumination test results,this paper used the active infrared CCD camera for video monitoring.A video smoke image processing and dust removal method was proposed as supplementary alarm verification.The method realized the automatic smoke recognition of the cargo compartment:firstly,the four-frame difference method was used to extract the moving target,and the median filtering was used to remove the random noise according to the slope value of the straight line fitted from the center of mass coordinates,combined with the characteristics of smoke and dust movement direction,SVM was used to distinguish smoke and dust to improve the accuracy of smoke identification.Real fire and dust interference experiments were conducted in the aircraft cabin,the results show that the video-based method for identifying aircraft cabin smoke can effectively remove the effects of image noise and dust,and provide early warning of smoke.