基于SVM的火焰图像特征融合算法研究

Research on Flame Feature Fusion Based on SVM

研究了支持向量机(SVM)的相关理论,并采集实验视频,在Matlab中进行了基于SVM的火焰图像特征融合算法仿真,结果表明,SVM分类器比贝叶斯分类器有更好的分类效果。

Firstly, theory of support vector machine （SVM） is studied. Secondly, experimental videos are collected, flame features are extracted, and flame feature fusion algorithm based on SVM is simulated in Matlab. Results show that the SVM classifier has better performance than Bayes classifier.