| 摘 要: 为了提高火焰检测的准确率和鲁棒性，提出了一种基于支持向量机(SVM)的火焰检测算法。首先根据火焰的颜色和运动特性，结合YCbCr颜色高斯模型和ViBe算法提取疑似火焰区域；为了提高检测的鲁棒性，并降低计算量，以秒为检测周期，提取疑似火焰区域的动、静态特征；最后将特征向量送入预训练好的SVM中进行最终判决。仿真实验表明，该算法具有较高的准确率，同时满足实时性要求。 |
| 关键词: 火焰检测；支持向量机；多特征融合 |
|
中图分类号: TP391
文献标识码: A
|
| 基金项目: 辽宁省科技厅科学事业公益研究基金项目“利用TRIZ理论构建辽宁省陶瓷矿产资源数据化系统”(项目编号：2016001005).辽宁省科技厅自然科学基金项目“基于多源探测的室内公共场所火灾预警关键技术研究”(项目编号：2014020116). |
|
| Flame Detection of Video Images Based on SVM |
|
ZHONG Ling,ZHANG Xingkun
|
(School of Information Science and Engineering, Shenyang University of Technology, Shenyang 110870, China )
|
| Abstract: In order to improve the accuracy and robustness of flame detection,the paper proposes a flame detection algorithm based on Support Vector Machine(SVM).Firstly,according to the color and motion characteristics of the flame,the algorithm extracts the suspected flame area with the YCbCr color Gauss model and the ViBe algorithm.Secondly,in order to improve the robustness of detection and reduce calculation,the algorithm extracts the dynamic and static characteristics of the suspected flame area on a one-second cycle.Finally the feature vector is put into the pre-trained SVM for detection.The simulation results show that the algorithm has high accuracy and meets the real-time requirements. |
| Keywords: flame detection;Support Vector Machine(SVM);multi-feature fusion |
