Abstract: The railway station data required by the computer-based interlocking automatic test software depends on manual input, and the degree of automation is not high. To solve this problem, this paper proposed an automatic extraction method of railway computer interlocking station data, took the verified interlocking host computer interface as the input of an image recognition system to analyze the display characteristics of signal equipment and define the data model. Based on OpenCVSharp computer vision library, the paper designed the recognition process by using the image recognition related algorithm to automatically extract the station data, developed the software using the above method to show the effect of signal equipment identification and data extraction. This method can accurately identify the signal, switch, and track circuit district in the interface of interlocking host computer, so then implement the automatic extraction of relevant station data.