基于最优阈值法的锅炉火焰图像处理的研究

Research on boiler flame image processing based on the optimal threshold method

工业锅炉炉膛内部是一个高温、高辐射、高灰度的环境,炉内燃料的燃烧过程伴随着剧烈而连续的发光发热的物理、化学反应。在锅炉火焰图像识别中,原始图像对该复杂环境变化非常敏感,并严重地影响图像的识别率。针对这一问题,本文介绍了一种提取锅炉火焰特征的最优阈值法。该方法依据统计学基本原理,对锅炉火焰自动识别的实际问题建立一个统计回归模型,并根据回归模型参数直接求取最优的分割阈值。

Industrial boiler firepot is an environment of high temperature, high radiation and high gray,The combustion process of fuel in firepot is a kind of intense and continuous physical and chemical reaction with light and heat. The original image is very sensitive to the complex environmental changes, and seriously affects the image recognition rate in the recognition of boiler flame image. To solve this problem, the paper introduces the optimal threshold method for extraction of boiler flame characteristics; the method based on statistics, and established a statistical regression model to the actual problem of automatic identification of boiler flame, and according to the regression model parameters calculates the optimal segmentation threshold directly.