In order to meet the requirements of the processing speed in real time of electro-optical images, a signal processing system based on multi-core DSP and FPGA is designed. And the hardware and software architecture design of this system is introduced. Images are transmitted by optical fibers and can be used to process CCD image data in different optical bands. Taking a fire monitoring system based on visible light CCD sensor as an example, utilizing the characteristics of flames and smoke during object combustion, and the image feature extraction algorithm is used to process the collected images in real time. After identifying the fire area, it is reported. Experimental results show that the signal processing system works reliably and stably, and has the ability of high-speed parallel image processing, which can simultaneously process multiple CCD images, achieve software online loading, image storage and unloading functions, and provide a universal solution for the system platform of electro-optical image processing.