An end-to-end alarm information identification method based on deep learning

Abstract

The invention discloses an end-to-end alarm information identification method based on deep learning, and the method comprises the following steps of intercepting a video frame image, and carrying outthe preprocessing of the image; cutting the image according to the line of the alarm information to obtain an image sheet containing each line of alarm information; performing text information detection on each image slice; and identifying the detected text information based on an end-to-end deep learning algorithm to obtain an alarm information identification result. The invention mainly aims atan intelligent auxiliary system for substation operation. When video image monitoring is carried out on multiple lines of dynamic alarm information displayed on a background electronic screen according to lines, according to the end-to-end identification method based on deep learning, all alarm information can be accurately identified, so that the alarm information on the electronic screen can beefficiently obtained, the information is timely fed back to the data analysis system, and the subsequent equipment can remove faults of a power transformation operation system according to the alarminformation.

Description

technical field [0001] The invention belongs to the technical field of computer vision, and in particular relates to an end-to-end alarm information recognition method based on deep learning. Background technique [0002] Alarm information recognition is actually to recognize the text information in the captured video images. Text recognition can input text into the computer system quickly and efficiently, realize human-computer interaction, reduce people's labor, and save manpower from boring and complicated work. Free up and improve processing efficiency. [0003] The traditional text recognition process is divided into several modules: character preprocessing, feature extraction, and pattern classification. Because text recognition is affected by many factors, such as complex background and noise, complex character structure itself, many types of characters, and many similar characters. In particular, when the traditional algorithm pre-processes characters, due to the sm...

AI Technical Summary

Problems solved by technology

[0005] The technical problem to be solved by the present invention is to provide an end-to-end alarm information recognition method based on deep learning to solve the problem that the existing text recognition algorithms based on deep learning are mostly used for the recognition of single sentences or phrases. The recognition effect is not ideal, and there are technical problems such as poor recognition accuracy

Images