Forest Fire Image Detection Method Based On Fpga Research And Implementation

The forest is the important resource in China, and it provides the important guarantee for thedevelopment of national economy in our country and the life of people. However, the forest firecauses the huge threat to forest resource, and brings the great loss to human ecology environmentand personal life property security. So, the forest fire has become a global concern problem, and theforest fire detection technology has attracted the worldwide attention. Therefore, the research offorest fire detection technology has the very important practice significance.With the development of science technology and computer network technology, a variety offorest fire image detection methods has risen. Firstly, building lots of watchtower is one of thedevelopmental directions of forest fire detection, but this forest fire detection technology is morebackward, and it has many shortcomings and insufficiencies. Secondly, setting up video monitoringsystem is the most mainline forest fire detection way adopted in the present domestic, and it is thestretch of traditional city monitoring detection, but it is easy to cause the omission. At the same time,the video monitoring is not the digital system, and many intelligent applications can not be realized.Last but not least, founding the intelligent forest fire detection system is the main developmenttrend of forest fire detection, and it realizes the intellectualization and informatization of forest firedetection.The task of this thesis is the research and implementation of forest fire image detectionmethods based on FPGA. The thesis mainly adopts FPGA to implement forest fire image detectionmethods, and implement the whole system design through using the D5M camera module, DE2development platform and VGA display. First of all, capture the image information through thefront of the camera. And then get through the video decoding module, the color space conversionmodule, the image processing module and fire detection module. Lastly, pass the video coding module, and display the last detection result through using VGA display. In the part of fire detectionmodule, the main forest fire detection way is image edge detection, and adopts the Sobel operator,the Prewitt operator and the Particle Swarm algorithm to implement the fire detection of forest fireimage, and well detect the part of flame of forest fire flame image. All the hardware realization ofthe whole system greatly improves the system processing speed, and has the high real-time. At thesame time, compare these methods in this thesis, and then analysis the quality of these methods,and supply the theory basis for the future research.