摘要
研究开发基于视频图像的嵌入式分析设备对林区火灾的早期预报、减少森林资源损失具有重要意义。以DSP为核心的嵌入式系统具有体积小、功能完备的优点,同时结合机器视觉领域的图像处理技术,使其成为基于视频图像的森林火灾监控系统的可行方案。本文在浙江省重大科技专项——基于太阳能的森林火灾监测技术与系统研制项目的支持下,在TI公司的DSP芯片TMS320DM642的基础上设计和开发了林火烟雾检测系统,以DM642为核心搭建系统硬件平台,基于实时可裁剪的DSP/BIOS内核设计了RF5系统软件框架,完成了系统多线程、多任务的实时调度。
烟雾是林区火灾发生早期的显著特征,对烟雾的及时检测是提高火灾预警能力的一个有效途径。根据开放环境下火灾烟雾检测的应用需求,提出了一种基于多特征融合以及神经网络/支持向量机判定的在线烟雾检测算法。该算法首先提出了适应开放环境且抗光照影响较好的烟雾颜色模型,并与Kalman运动检测图像进行融合得到疑似烟雾区域;接着利用小波变换等方法对疑似区域提取烟雾的轮廓不规则性、面积扩散性、烟雾区域模糊性以及烟雾边缘的低频振荡性四个时空域特征,作为离线训练完成的BP神经网络/支持向量机SVM分类器的四个输入;最后根据分类器的输出来判定所监测的林区是否发生火情。
本系统通过现场调试和实验表明,系统稳定可靠,有较高的实效性,系统设计和资源分配正确合理;本文烟雾检测算法能够实时进行视频烟雾检测,对开放环境较为鲁棒,能较好的满足开放环境下火灾烟雾的监测预警,基本达到了预期效果。
It is of great significance to take a research and develop a system of video image analysis for early warning of forest fires and reduce the loss of forest resources based on embedded technology. Because of its small volume and fully functional, the embedded system taking DSP as the core are adopted .Combining with the technology of image processing in machine vision field, it’s also becoming a good choice for the forest fire monitoring system. The paper developed the fire smoke detection system based on TMS320DM642 made by TI, took DM642 as the core kernel, and planned the RF5 software framework by DSP/BIOS to complete the multitasking real-time scheduling by the support of the projects of major science and technology in Zhejiang -- solar energy-based forest fire monitoring technology and system development.
Smoke is a remarkable characteristic for early forest fire, the timely detection of smoke is an effective approach to improve the ability of fire warning. According to actual application requirements of smoke detection in the open environment, an on- line algorithm based on multi-feature integration and neural network or Support Vector Machine (SVM) was proposed. Firstly, a smoke-color model adapted to open environment and resistance to light was presented, then, got the suspected smoke area with Kalman motion detection by image fusion. Secondly, four spatial characteristics of smoke were extracted as follows: contour irregularities, area of avian, ambiguity of smoke region, low frequency oscillation of smoke use wavelet and so on, as the inputs of BP neural network and support vector machine trained completed offline. Finally, the paper used the classifier’s outputs to determine whether there was occurring fires.
The debugging and experiments’results show that the system is reliable and higher effective, the design of system and resource allocations are correctly reasonable. The algorithm of paper can realize real-time video smoke detection under certain conditions, and adapts to the open environment, meets the requirements of fire smoke warning in open environment and desired effects.
引文
[1].黄普希,张昊.国外视频烟雾探测技术简析[J].论文荟萃.2007.1(6):13-16.
[2].杨猛,赵春晖,潘泉,张绍武.基于小波分析的烟雾多特征融合和空间精度补偿森林火情检测算法[J].中国图像图形学报.2009.14(4):694-700.
[3].帅师,周平,汪亚明,周维达.基于小波的实时烟雾检测[J].计算机应用研究. 2007.24(3):309-311.
[4].林宏,陈旭,强振平.视频火灾自动预警中烟雾检测的研究[J].人工智能及识别技术.2009.2:81-83.
[5].练秋生,李党.融合多种特征的烟雾图像检测算法[J].光学技术.2009.35(4):523-529.
[6].王欣刚,魏峥,刘东昌,郑晓隆.基于烟雾动态特征分析的实时火灾检测[J].计算机技术与发展.2008.18(11):9-12.
[7].邓彬,刘辉,连国云,陈静.基于视频的烟雾检测[J].长沙大学学报.2007.21(5):87-89.
[8].夏文德,练秋生.基于视频处理的烟雾检测快速算法[J].电子技术. 2007.34(11):135-138.
[9].Dong Keun Kim,Yuan-Fang Wang. Smoke Detection in Video, in 2009 World Congress on Computer Science and Information Engineering[C].vol.5, pp: 759-763.2009.
[10].Hidenori Maruta, Yasuharu Kato, Akihiro Nakamura and Fujio Kurokawa. Smoke detection in open areas using its texture features and time series properties, in IEEE International Symposium on Industrial Electronics (ISIE 2009), pp: 1904-1908.2009.
[11].Thou-Ho (Chao-Ho) Chen, Yen-Hui Yin, Shi-Feng Huang and Yan-Ting Ye. The Smoke Detection for Early Fire-Alarming System Base on Video Processing, in Proceedings of the 2006 International Conference on Intelligent Information Hiding and Multimedia Signal Processing[C],pp.427-430, 2006.
[12].Paolo Piccinini, Simone Calderara, Rita Cucchiara. RELIABLE SMOKE DETECTION SYSTEM IN THE DOMAINS OF IMAGE ENERGY AND COLOR, in Image Processing, 2008(ICIP 2008). 15th IEEE International Conference, pp: 1376-1379.2008.
[13].Xiaohu Lu,Hong Zhang. Adaptive color claddification with gaussian mixture model. Proceedings of Robotics and Biomimetics (ROBIO). 2005 IEEE International Conference[C].2005:330-335.
[14].王青松.DSP低功耗设计技术[D].西安电子科技大学硕士学位论文.2007年.
[15].Stauffer C , Grimson WEL. Adaptive background mixture models for real-time tracking [ R ] In Proc. IEEE Conference on Computer Vision and Pattern Recognition , IEEE Computer Society , 1999 :246-252.
[16].严勇,黄席樾,刘爱君. Kalman滤波在运动图像背景提取及更新中的应用[J].自动化与仪器仪表.2006.2:28-30.
[17].Gonzalez R C.Digital Image Processing [ M ] ( 2nd ) .Beijing: Publishing Hourse of Electronics Industry, 2003:461-475.
[18].张海涛,蔡文寰,董有尔.基于DM642的图像处理系统设计及应用[J].现代电子技术.2008.12:125-127.
[19].胡爱闽.基于DM642的网络视频监控系统[J].华东交通大学学报.2008.25(4):47-50.
[20].Texas Instruments.TMS320C6000 DSP External Memory Interface (EMIF) Reference Guide[M].www.ti.com,2006.
[21].TEXAS INSTRUMENTS INCORPORATED. TL16C752B. 3.3-V Dual UART with64-Byte FIFO [ EB /OL ].2006.http:// www.Ti.com.
[22].Texas Instruments.TMS320DM642 Video/Imaging Fixed-Point Digital Signal Processor[M].www.ti.com,2002.
[23].Texas Instruments.The TMS320DM642 Video Port Mini-Driver for TVP5146 and TVP5150 decoder[M].www.ti.com,2007.
[24].Texas Instruments.Reference Frameworks for eXpressDSP Software:RF5,An Extensive,High-Density System[M].www.ti.com,2004.
[25].杨尚罡,黄席樾,李宝林,张伟.基于DSP/BIOS的隧道视频监控系统[J].重庆大学学报(自然科学版),2006,29(7):81-84.
[26].程翔宇,李迪,冯寿廷,王小良.基于DSP/BIOS的运动控制系统设计[J].机床与液压,2007,35(5):157-159.
[27].Texas Instruments.TMS320 DSP/BIOS User's Guide[M],2004.
[28].TEXAS INSTRUMENTS INCORPORATED. TMS320C6000 DSP/BIOS API Refenerce guide ( Rev. F) [ EB /OL ].2006 . http://www. Ti. com.
[29].Texas Instruments.The TMS320DM642 Video Port Mini-Driver for TVP5146 and TVP5150 decoder[M].www.ti.com,2007.
[30].Texas Instruments.DSP/BIOS Driver Developer's Guide[M],www.ti.com,2002.
[31].周一君.基于DM642的嵌入式林区火焰监控系统研究[D].杭州电子科技大学硕士学位论文.2009年.
[32].单潮龙,马伟明,贲可荣等.BP人工神经网络的应用及其实现技术[J].海军工程大学学报.2000.4:16-22.
[33].李晓歌,邵丽红,李娜.人工神经网络BP算法的实现[J].人工智能及识别技术.2007.12:921-923.
[34].杨期鹤,曹年红,陆信人.基于DSP的BP神经网络的硬件实现[J]信号理.1998.14(2):183-188.
[35].梅建新,段汕,秦前清.基于支持向量机的特定目标检测方法[J]武汉大学学报(信息科学版).2004.29(10):912-915.
[36].冼广铭,曾碧卿,冼广淋.支持向量机在分类和回归中的应用研究[J].计算机工程与应用.2008,44(27):134-136.
[37].张健沛,徐华.支持向量机(SVM)主动学习方法研究与应用[J]计算机应用.2004.24(1):1-3.
[38].闫建强,马莉.基于DSP/BIOS的视频监控系统多线程调度研究[J].机电工程.2010.27(9):32-35.
[39].Feiniu Yuan. A fast accumulative motion orientation model based on integral image for video smoke detection, in Pattern Recognition Letters 29 (2008), pp: 925-932, 2008.
[40].Jerome Vicente, Philippe Guillemant. An image processing technique for automatically detecting forest fire, in International Journal of Thermal Sciences 41 (2002),pp. 1113-1120, 2002.
[41].v.Cappelllnl, L. Mattil, A. Mecoccl. AN INTELLIGENT SYSTEM FOR AUTOMATIC FIRE DETECTIONIN FORESTS, in Image Processing and its Applications, 1989.Third International Conference[C],pp.563-570, 1989.
[42].WANG Tang, CHEN Gong, ZHOU Heqin. A novel background modeling approach for accurate and real-time motion segmentation, in Signal Processing (ICSP2006), 8th International Conference[C]. Vol.2.2006.
[43].Zhengguang Xu, Jialin Xu. Automatic Fire smoke Detection Based on Image Visual Features,in 2007 International Conference on Computational Intelligence and Security Workshops[C],pp.316-319.2007.
[44].Chao-Ching Ho, Tzu-Hsin Kuo. Real-Time Video-Based Fire Smoke Detection System, in 2009 IEEE/ASME International Conference on Advanced Intelligent Mechatronics[C], pp.1845-1850.2009.
[45].Nobuyuki, Fujiwara, Kenji Terada. Extraction of a Smoke Region Using Fractal Cording [A]. International Symposium on Communi cations and Information Technologies[C]. 2004.
[46].B Ugur Toreyin, Yigithan Dedeoglu,A Enis Cetin. Wavelet Based Real-Time Smoke Detection in Video [A] . 13th European Signal Processing Conference[C]. 2005.
[47].Xueming Shu, Hongyong Yuan,Guofeng Su , et al. A new method of laser sheet imaging-based fire smoke detection [J]. Journal of Fire Sciences, 2006, 24 (2): 95 -104.
[48].Khananian A , Fraser R H , Cihlar J . Automatic detection of fire smoke using artificial neural networks and threshold approaches applied to AVHRR imagery [J] . IEEE Transactions on Geoscience and Remote Sensing, 2001, 39 (9): 1859 -1870.
[49].Searle, Stephen J. Background modeling and target segmentation via modified Kalman filtering [A]. in Proc of SPIE[C] . 2004.
[50].Chan W L , Choi H , Baraniuk R. Coherent multiscale image processing using dual-tree quaternion wavelets[J ].IEEE Transactions on Image Processing,2008,17 (7):1069-1082.
[51].Collins R T , Lipton A J , Kanade T. A System for Video Surveillance and Monitoring[ C].Proc. of American Nuclear Society 8th Int . Topical Meeting on Robotics and Remote Systems.1999.
[52].Texas Instruments.DSP/BIOS Driver Developer's Guide[M],www.ti.com,2002.
[53].Texas Instruments.TMS320C6000 Chip Support Library API Reference Guide[M].www.ti.com,2005.
[54].黄普希,张昊.国外视频烟雾探测技术简析[J].论文荟萃.2007.1(6):13-16.
[55].程晓舫,王瑞芳,张维农等.火灾探测的原理与方法(上)[J].中国安全科学学报,1999,9(1):24-29.
[56].袁非牛,张永明,刘士兴等.基于累积量和主运动方向的视频烟雾检测方法[J].中国图象图形学报.2008.13(4):808-812.