首页> 外国专利> VIDEO BASED SMOKE DETECTION METHOD ROBUST TO LIGHT CONDITIONS

VIDEO BASED SMOKE DETECTION METHOD ROBUST TO LIGHT CONDITIONS

机译:基于视频的烟雾检测方法在光照条件下非常稳定

摘要

PURPOSE: An image based smoke detection method for changes of brightness and chroma of a photographing object is provided to reduce the domain of false detection. CONSTITUTION: Pixels exceeding the threshold of predetermined brightness and chroma are extracted to a candidate area. A statistical property of Gaussian mixture modeling function about the candidate area is obtained by using information for the brightness and chroma of the pixels comprising the candidate area. Whether the candidate area has fire is determined by the statistical property of Gaussian mixture modeling function about the candidate area. [Reference numerals] (AA) Photographing an image; (BB) Current image; (CC) Extracting a background image; (DD) Difference-image; (EE) Measuring the brightness and chroma of the current image, and setting the threshold of predetermined brightness and chroma classified by four steps; (FF) Verifying the excess of each pixel forming the difference-image over the predetermined brightness and chroma; (GG) Extracting a candidate area; (HH) Gaussian mixture modeling; (II) Comparing with smoke features; (JJ) Detecting smoke; (KK) Alerting
机译:目的:提供一种基于图像的烟雾检测方法,用于改变拍摄对象的亮度和色度,以减少错误检测的范围。组成:超过预定亮度和色度阈值的像素被提取到候选区域。通过使用关于构成候选区域的像素的亮度和色度的信息,获得关于候选区域的高斯混合建模函数的统计特性。候选区域是否着火取决于候选区域的高斯混合建模函数的统计特性。 [附图标记](AA)拍摄图像; (BB)当前图像; (CC)提取背景图像; (DD)差异图片; (EE)测量当前图像的亮度和色度,并设置按四个步骤分类的预定亮度和色度的阈值; (FF)验证形成差异图像的每个像素在预定亮度和色度上的剩余量; (GG)提取候选区域; (HH)高斯混合建模; (二)与烟雾特征比较; (JJ)检测烟雾; (KK)警报

著录项

相似文献

  • 专利
  • 外文文献
  • 中文文献
获取专利

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号