<p>This study attempts to address the technical challenges of early fire detection and rapid remote verification in unattended locations. A fire image transmission method based on the data bus of Automatic Fire Alarm System (AFAS) is proposed. The Gaussian difference algorithm is used to extract feature parameters of background image and fire image. Difference regions are accurately identified by comparing the feature extrema of these two images. An encoding strategy is adopted to perform compression encoding only on the difference regions instead of the full fire image. This significantly reduces the transmitted data volume and improves the transmission efficiency, enabling remote transmission of fire image via the existing data bus of AFAS. Experimental results show that the remote transmission time of fire image is mainly affected by the resolution of the image acquisition module, the proportion of the difference region, and the transmission distance. By adjusting the resolution of the acquisition module and the layout position of the detector, the effective transmission distance could be extended. The proposed method requires neither modifying the wiring of the existing AFAS nor building a new dedicated image transmission line, thus offering significant application advantages.</p>