论文部分内容阅读
针对炉膛火焰图像监测系统存在的灭火判别问题,提出了一种炉膛火焰灭火判别方法。该方法通过对火焰图像的分析,提取了用于灭火判别的两个特征值,然后对这个特征空间使用支持向量机进行识别分类,结果表明特征量提取是成功的。该判别方法能够正确对火焰图像进行灭火判别,正确率很高,据此可知支持向量机方法用于灭火判别是可行的。
Aiming at the problem of fire extinguishing discrimination existing in the furnace flame image monitoring system, a fire extinguishing discriminating method for furnace hearth is proposed. The method extracts two eigenvalues used for fire extinguishing discrimination by analyzing the flame image, and then uses SVM to classify the feature space. The result shows that the feature extraction is successful. The discriminant method can correctly discriminate the flame image and the correctness rate is very high. From this we can see that the SVM method is feasible for fire extinguishing discrimination.