节点文献

基于三角扩散模型的遮挡林火图像中火源点确定

Determination of Fire Points under Occlusion in A Triangular Diffusion Model

  • 推荐 CAJ下载
  • PDF下载
  • 不支持迅雷等下载工具,请取消加速工具后下载。

【作者】 赵齐韩宁罗琴娟

【Author】 ZHAO Qi;HAN Ning;LUO Qin-juan;School of Technology,Beijing Forestry University;

【机构】 北京林业大学工学院

【摘要】 针对目前森林火灾数字视频监控系统中火源点被遮挡情况下无法定位的问题,提出了一种基于林火烟雾三角扩散模型的方法,用于确定林火报警图像上被遮挡火源点的像素坐标。该方法首先对已被识别的林火烟雾图像进行分割和边缘提取,进而运用光流法分析得到林火烟雾运动趋势,在此基础上构建三角扩散模型,从而估算出图像中遮挡情况下的林火火源点像素坐标。实验结果表明,在烟雾遮挡但能够呈现出三角轮廓时,其像素误差在20个像素点内,这一结论可为实现林火数字视频监控系统的全视域定位提供理论支持。

【Abstract】 In order to solve the problem of the location of fire points under occlusion in a digital forest fire video monitoring system,a method is presented based on triangular diffusion model forest fire smoke to determine the pixel coordinates of fire points on a forest fire alarm image under occlusion.First,this method involves segmentation and edge extraction of an identified forest fire smoke image,followed by forest fire smoke movement trend obtained through analysis with an optical flow method,based on which a triangular diffusion model is constructed,based on which the pixel coordinates of fire points on a forest fire alarm image under occlusion are estimated.The experimental result shows that smoke occlusion can reveal a triangular contour,the pixel error is within 20 pixel points,and this conclusion can provide theoretical support for realizing the full-view positioning of a forest fire digital video monitoring system.

  • 【文献出处】 林业机械与木工设备 ,Forestry Machinery & Woodworking Equipment , 编辑部邮箱 ,2016年09期
  • 【分类号】S762.3
  • 【被引频次】1
  • 【下载频次】76
节点文献中: 

本文链接的文献网络图示:

本文的引文网络