A Novel Embedded System Architecture for Fire Detection Based on Image Processing

This paper presents novel logic/software co-work architecture for embedded high definition image processing platform, which is built by the considerations of system level, board hardware level, and the tasks partition between CPU processing and programmable logic based on the latest launched System on Chip Field Programmable Gate Array (Soc FPGA) – Xilinx ZC7020. For this case, we comprehensive analyze of the critical data paths: the uniform Advanced Extensible Interface (AXI) processing between processing system (PS) and processing logic (PL), including high definition video pass through PL to PS and PS software processing send to PL for speed up. We have included the transplant of opensource Linux, multiprocessing cooperative control and boot loader in PS side. Since the general platform is proposed, a fire detection approach based on high definition image processing is implemented. Experiment results indicated the feasibility and universality of the embedded system architecture.

[1]  Jialin Wang,et al.  A SVM approach for vessel fire detection based on image processing , 2012, 2012 Proceedings of International Conference on Modelling, Identification and Control.

[2]  Hasan Demirel,et al.  Fire detection in video sequences using a generic color model , 2009 .

[3]  Turgay Çelik,et al.  Fire Detection in Video Sequences Using Statistical Color Model , 2006, 2006 IEEE International Conference on Acoustics Speech and Signal Processing Proceedings.

[4]  Jong-Myon Kim,et al.  Fire flame detection in video sequences using multi-stage pattern recognition techniques , 2012, Eng. Appl. Artif. Intell..

[5]  Turgay Çelik,et al.  Automatic fire detection in video sequences , 2006, 2006 14th European Signal Processing Conference.

[6]  Wen-Bing Horng,et al.  A new image-based real-time flame detection method using color analysis , 2005, Proceedings. 2005 IEEE Networking, Sensing and Control, 2005..