锅炉火焰动态跟踪识别算法的研究

A DYNAMIC METHOD USED TO RECOGNISE FLAME BURNING STATUS IN THE BOILER

采用动态跟踪火焰焰芯算法和改进的一维灰度直方图算法用于燃烧控制 ,很好地解决了实时监测锅炉火焰的燃烧状态和品质问题 .10 0 0幅典型图像的仿真结果 ,同目测判别结果比较 ,此方法识别率优于 99.8% .

A dynamic tracing flame center algorithm and an advanced two-dimensional gray level histogram thresholding algorithm were adopted for the combustion contral. It solves the problem of real-time inspecting the burning state and quality of boiler flame effectively. Comparing the simulative experimental results with artificial judgement on one thousand representative images, the recognition efficiency of the method is more than 99.8 percent.