Research Of Embedded Smoke Identification System Based On OpenCV

With the more and more perfect development of computer and technology,accurate detection and tracking employing computer image technology based on smoke characteristics,color,movement and texture characteristics has become a hot research topic in china.This paper presents a scheme of smoke recognition to achieve intelligent recognition smoke,by using the Linux operating system with Coretext-A8 platform based on embedded technology,graphics technology support,recognition of the acquisition and processing of video encoding,and OpenCV technology in embedded system.First of all,the background and significance of the research was introduced in this paper.,as well as the route method and the key technology of the smoke recognition system were put forward,according to the smoke detection research at home and abroad.Secondly,the related technology,such as UBUNTU12.0.4,OpenCV2.4.3 algorithm and wavelet technology,etc,were analysised soas to lay a theoretical foundation for the research topic.Then,the algorithm of image recognition and video tracking was analyzed with the characteristics,accoding to the image characteristics,video processing,motion target recognition and tracking.On the basis of the above research and analysis,this paper proposed a OpenCV video smoke detection system based on the analysis algorithm accoding to the hardware and soft environment required for development system.Finally,this paper implemented the cross compile of the depended library,and achieved the transplant of OpenCV 2.4.3 based on Coretext-A8 kernel well.Among diverse types of mobile robots,spherical robots have become increasingly attractive in recent yearts.Spherical robot has a ball-shaped outer shell to accommodate the whole mechanism,including control devices and energy sources.They are believed to have several advantages,such as low friction,low-energy consumption,small size and space,good sealing,good dustproof and shockproof,good stability omni-direction moving.These advantages provide the spherical robots with better developing potential compareing with the traditional mobile robots,such as wheeled,legged and tracked ones.Therefore,the spherical robots will have good perspective applications.They can be used as traffic conveyance,playing toys,or be used in the fields of surveillance,reconnaissance,space exploration and entertainments,even can be used as intelligent bombs in the field of military affairs.In this paper,embedded smoke detection system platform is built based on OpenCV.Based on the platform,the static image and dynamic smoke detection experiments were carried out.Through the analysis and comparison of the experimental results,it is proved that the system is not only feasible and efficient,but also stable and reliable.Therefore,it is effective to use this system to detect smoke.This study not only has far-reaching significance and application value in forest fire,moke detection and recognition,but also has the important application value and broad market prospects for the development of intelligent transportation,smart security,intelligence,public safety and other Home Furnishing.