Image Recognition And Processing Of Host Computer Interface Of Computer-Based Interlocking

The host computer interface of railway computer-based interlocking is the main basis for the station attendant to issue route control commands and master the status of signal equipment in real time.The interface not only reflects the topological relationship of station signal equipment layout and the real-time status of outdoor signal equipment,but also reflects the setting and layout of operation buttons.At present,the testing process of interlocking software still needs manual participation.For large stations,the workload is huge and error prone,and there are some deficiencies such as low efficiency and long time-consuming.Therefore,the third-party interlocking software testing concept came into being,and how to automatically extract the attribute data of signal equipment required by the third-party interlocking software testing system has become an urgent problem to be solved.By studying the relevant technical specifications of computer-base interlocking operation display,this thesis summarizes and analyzes the drawing characteristics of each signal equipment in the interlocking host computer interface,introduces Open CVSharp image recognition technology,and combined with the existing computer-based interlocking simulation system,designs and develops the interlocking host computer interface recognition software using c# language under the environment of visual studio 2017.The software takes the interface image of the interlocking host computer as the input,and realizes the high-quality and fast output of the attribute data of the signal equipment through image recognition,so as to provide the static and dynamic data of the signal equipment for the automatic test system of the third-party computer-based interlocking software,which is helpful to improve the universality and intelligent level of the third-party interlocking software test system.The main research work of this thesis is as follows:1.Introduce the computer-based interlocking system,analyze the display characteristics of signal equipment in the interlocking host computer interface in combination with relevant technical specifications,clarify the functional requirements of the interlocking host computer interface identification software,and design the overall software architecture and software workflow.2.Study the relevant algorithm principle of image recognition,clarify the interface image recognition processing flow of interlocking host computer,determine the key technologies required for interface image recognition,and analyze,introduce and display the application effect.3.According to the overall design architecture of the software,the functional modules of the interlocking host computer interface identification software are designed in detail to realize the identification and positioning of the signal,switch,track circuit district,rail insulation,signal button and function button display area,define the content and format of relevant attribute data,and realize the detection and extraction of attribute data.4.Realize the development of interlocking host computer interface identification software,show the functions realized by the software according to the software workflow,and provide the attribute data of signal equipment to the computer-based interlocking automatic test system to realize the automatic generation of station topology diagram and interlocking table.This thesis focuses on the application of image recognition technology to the interlocking host computer interface to realize the identification of station signal equipment,automatically extract the static and dynamic attribute data related to each signal equipment,provide the required signal equipment data for the computer-based interlocking automatic test system,and help to improve the automation level of interlocking software test.