| 本文已被：浏览 3417次 下载 5839次 |
码上扫一扫！ |
|
|
| 采用圆周对齐边缘方向直方图的视频烟雾检测方法 |
|
袁非牛1,2, 方志军1,2, 杨勇1,2, 方玉明1,2, 杨寿渊1,2
|
|
1.江西财经大学 信息管理学院, 江西 南昌 330032;2.江西省公共安全视频技术研究中心(江西财经大学), 江西 南昌 330032
|
|
| 摘要: |
| 视频烟雾检测具有响应速度快、非接触等优点,但由于烟雾形状、色彩、纹理千差万别,目前的算法很难取得令人满意的检测效果.为此,提出了一种鲁棒的特征提取方法,采用支持向量机(support vector machine,简称SVM)进行检测.首先,提取边缘方向直方图(edge orientation histogram,简称EOH).然后,采用圆周平移方式将EOH的最高柱变换到EOH的固定位置,消除了旋转变换的影响.为了进一步增强特征的鲁棒性,提取图像亮度和饱和度分量的Hu不变矩、均值、偏差、偏度和峰度特征.最后,将这些特征组成一个38维的特征矢量,采用SVM训练和识别烟雾.实验结果表明,这些特征具有很好的分类性能,能够在较大的训练库和测试库上达到98%和85%以上的检 测率. |
| 关键词: 边缘方向直方图 圆周对齐 Hu不变矩 支持向量机 |
| DOI： |
| 分类号: |
| 基金项目:国家自然科学基金(61063034,61363038);江西省高校科技落地计划(KJLD12066);江西省青年科学家培养对象资助项目(20142BCB23014) |
|
| Video Smoke Detection Based on Circularly Aligned Edge Orientation Histogram |
|
YUAN Fei-Niu1,2, FANG Zhi-Jun1,2, YANG Yong1,2, FANG Yu-Ming1,2, YANG Shou-Yuan1,2
|
|
1.School of Information Technology, Jiangxi University of Finance and Economics, Nanchang 330032, China;2.Research Center for Jiangxi Public Security Video Technology (Jiangxi University of Finance and Economics), Nanchang 330032, China
|
| Abstract: |
| Video smoke detection has many advantages such as fast response and non-contact. Due to large variance of smoke shape, color and texture, it's difficult for existing methods to achieve satisfactory results. This paper proposes a robust feature extraction method by using support vector machine (SVM) for classification. First, an edge orientation histogram (EOH) is extracted. Then, circular shift technique is used to transform the maximum value bin of EOH to the fixed bin of EOH, thus eliminating the influence of rotation. To further enhance the robustness of features, Hu invariant moments, mean, deviation, skewness, and kurtosis are extracted from both illuminance and saturation component images converted from original RGB images. Finally, all the features are combined together to form a 38-dimentional feature vector, and SVM is used to train and classify smoke images. Experiments show that the features have good discrimination capabilities, and the method can achieve about 98% and 85% detection rates on selected large training and testing data sets. |
| Key words: edge orientation histogram (EOH) circular alignment Hu invariant moment support vector machine (SVM) |
