Flame Monitoring System Based On The Dm642 Embedded Forest Areas For Study

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. The objective of this paper is to develop and design a surveillance system for flame detections. Main work of this thesis consists of:(1)building and designing a hardware platform of video flame monitoring based on TMS320DM642.(2)A advanced application of reference frame(RF5) based on DSP/BIOS has been utilized to collect, process and display images.(3)A flame detection algorithm based on fuzzy logic is proposed. In order to improve the precision of fuzzy control, the membership function under fuzzy sets (including input and output) is then optimized off-line by genetic algorithm, finally implemented for video based fire alarming. (4)According to the requirements of a real-time system, the rate of CPU utilization is greatly improved by using Ping-pong buffering (implementing parallel operation of EDMA data transfer and CPU image processing) and coding optimization.After debugging the hardware and software together of the system and experiments on field,experimental results and data show that the video flame detection can be accurately and timely performed by the system under some specified cases. And preset research goals have been achieved.