融合多特征的视频火焰检测  

Video Flame Detection Based on Fusion of Multi-feature

作　　者：严云洋[1,2] 杜静[1,2] 高尚兵[1] 周静波[1] 刘以安[2] 

机构地区：[1]淮阴工学院计算机工程学院,淮安223003 [2]江南大学物联网工程学院,无锡214122

出　　处：《计算机辅助设计与图形学学报》2015年第3期433-440,共8页Journal of Computer-Aided Design & Computer Graphics

基　　金：国家自然科学基金(61402192);教育部科学技术研究重大项目(311024);江苏省"六大人才高峰"项目(2013DZXX-023);江苏省"333工程";江苏省"青蓝工程";淮安市"533工程";江苏省高校自然科学研究重大项目(11KJA460001);淮安市科技计划项目(HAG2013057;HAG2013059)

摘　　要：对于复杂的场景,人类视觉系统选择性注意机制能够不需要训练而快速地定位到图像中的显著目标上.文中结合火焰的先验信息,基于显著性的四元数离散余弦变换算法来检测视频中的火焰.首先根据火焰在RGB空间中3个颜色分量之间的特殊关系改进了2个火焰颜色特征公式,得到2幅火焰颜色的特征图;然后通过计算疑似火焰区域的LBP特征向量的距离得到火焰的纹理特征图;再根据火焰内部的动态纹理、火焰闪烁频率特征计算改进后的火焰高频过零次数,得到火焰的动态特征图;最后将这4幅火焰特征图构成一个四元数,利用四元数离散余弦变换得到最终的火焰显著图.在Bilkent大学的火焰视频库中进行实验的结果表明,该方法具有准确率高、鲁棒性强的特点,优于对比的其他视频火焰检测算法.The selective attention mechanism of human visual system can quickly locate salient targets in the image without training in complex scenes. Combined with the prior information of flame, the saliency based algorithm of quaternion discrete cosine transform is proposed to detect the flame in video. Firstly, two color feature maps were obtained by improving the process of flame color feature extraction according to the special relationship between R,G and B of the flame color in the RGB space. Secondly, the texture feature map was generated by computing the distance of LBP feature vectors. Thirdly, the motion feature map was provided by calculating the zero-crossing frequency according to the dynamic texture characteristics and flicker frequency of the flame. Finally, the four feature maps are regarded as all components of a quaternion and the flame saliency map is produced by quaternion discrete cosine transform with the four feature maps. The experimental results on the flame video library of Bilkent university show that the proposed method is obviously superior to the corresponding algorithms on the accuracy and robustness.

关 键 词：视觉显著性分析 离散余弦变换 四元数 视频检测 火焰检测 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]