Research On Database Operation Recognition Technology Based On Deep Learning

The security of power enterprise information system database is very important.The operation and maintenance audit system connects power enterprise information systems through single sign on,and records and audits the operation and maintenance process,so as to trace the operation of operation and maintenance personnel in the operation and maintenance database process,and realize audit and supervision.However,log audit and artificial post audit are prone to omission,and can not prevent illegal operation in time,so there are security risks in the operation and maintenance process.It is an effective way to realize the automation of operation and maintenance audit by using artificial intelligence image recognition technology to build the identification and audit model for illegal operation and maintenance.In this paper,the application technology of artificial intelligence image recognition is used,and the image recognition method based on deep learning is used to construct the image audit model for the audit of database operation behavior.In this paper,Python+OpenCV+YOLO v5 solution is used to identify and detect the target of the database operation area dialog box of Navicat.At the same time,CTPN+YOLO v5+CRNN method is used to recognize the OCR image of the operation instruction environment line,and then the contents of the operation instruction of database operation and maintenance are comprehensively analyzed.The development of background audit system is based on the implementation of audit model,which provides users with multiple functional modules including rule management,policy management,work order management and early warning management to achieve efficient and dynamic operation and maintenance audit management.To provide users with richer database operation and maintenance rules.strategy formulation and implementation,improve operation and maintenance audit ability and enrich audit means and content.At the same time,it supports blocking illegal operations and automatically recording audit content,which is convenient for subsequent audit verification.