Abstract: Dispatching commands are the core instructions of railway transportation dispatching and command. Accurately and quickly conveying dispatching commands is a prerequisite for ensuring the safe, efficient and orderly operation of the railway system in a complex operating environment. At present, there are many scenarios in the process of railway transportation production where written dispatching commands are conveyed. It is required to read the written dispatching command information manually and complete the related data entry work, which is time-consuming, labor-intensive, inefficient, and prone to problems such as incorrect information entry and missed transmission. Based on the characteristics of written dispatching commands, this article studies the intelligent recognition and extraction methods of written dispatching command information. It combines image segmentation processing with the classic OCR algorithm to enhance the recognition ability of table structure and text content, and more accurately segment, locate and recognize the text data in written dispatching commands. Combined with the dispatching command template and language norms, the extraction of key information included in dispatching command is realized. The experimental results show that performing table segmentation processing before character recognition has a significant effect on improving the accuracy of text recognition of dispatching commands and is convenient for the subsequent automatic extraction of structured data of dispatching commands.