The platform safety gate is important safety protection equipment in urban rail transit, which makes the rail area relatively independent from the platform waiting area, ensures the safety of passengers, reduces the noise pollution brought by the subway train to the platform, and provides a comfortable waiting environment for passengers. In order to solve the problems of low intelligent degree and single debugging method of railway station safety door equipment data monitoring, a correction algorithm using machine learning based on image grid is proposed. Firstly, based on virtual instrument technology, a set of acoustic signal acquisition and processing systems for sound field visualization is designed and implemented. Then, based on the analysis of requirements, the hardware configuration and system software design are carried out. Finally, the extraction technology of image feature information is adopted, which can reduce the operation time of image target recognition and make the security door control system have real time. The experimental results show that the calibration algorithm is used to calculate the coordinate values of the actual road by using the third-order fitting method. Compared with the coordinate values of the standard grid, the average error of X is 0.0662%, and the average error of Y is 0.0011%. It can not only improve the accuracy of judgment, but also meet the real-time requirements of video monitoring. The system can realize wireless monitoring on the status of platform safety door equipment using machine learning, improve the efficiency of subway operation and the flexibility of station staff maintenance and protection, and ensure the safety and reliability of the platform safety door system.
