摘要
图像型火灾探测的核心问题是火焰和疑似火焰物体的分类和识别。以火灾视频和疑似火灾视频为分析对象,提取了火灾图像的面积重叠率、圆形度以及火焰尖角数目三个特征量,选择快速支持向量机进行分类器训练,最终利用训练好的分类器实现了火焰及干扰物体的分类识别问题。实验结果表明,该算法提高了火灾图像的分类精度和火灾识别的准确率,同时具有较高的检测效率。
The key problem of image fire detection is the recognition and classfication for fire flame and suspected fire object. Firstly extracted the three features of image based on fire video and suspected fire video,which were variance ratio of flame areas,circularity and the number of sharp angles. Then trained the classifier with FCSVM. At last,classified the fire and suspected fire object with the trained classifier. The experiment results show that the algorithm improves the classification precision of fire image and accuracy rate of fire recognization,and the algorithm has higher detection efficiency.
出处
《计算机应用研究》
CSCD
北大核心
2010年第10期3985-3987,共3页
Application Research of Computers
基金
国家自然科学基金资助项目(60634030)
高等学校博士学科点专项科研基金(20060699032)
陕西省教育厅专项基金资助项目(08JK319)
关键词
快速支持向量机
视频
火灾探测
FCSVM（ fast classification for support vector machine）
video
fire detection
