In order to solve the problem of incomplete preservation of the texture details of the melted image,this research builds a long-distance vision system based on the multi-source information fusion technology of infrared and visible light,and extracts and integrates the target and scene image data of the multi-source image to make the image more complete and clear.The infrared image acquisition module uses the FPGA main control chip model EP4CE40F23C8 to generate timing control signals to provide digital signals such as clock signals and integral signals that work normally,and an impedance conversion circuit is designed to buffer the detector output signals.An attention module is added to the generator,and the visible light image and the infrared image are encoded to extract their respective image features.The experimental results show that the fusion image of the long-distance vision system in this study has rich detailed information,the picture is clear and the contrast is strong,and the peak signal-to-noise of the fusion image is above 35 dB.