Character identification method based on deep characteristic and shallow characteristic fusion, and storage medium

Abstract

The invention provides a character identification method based on deep characteristic and shallow characteristic fusion, and a storage medium. A convolutional neural network model is trained to detectthe deep characteristics of an image to be identified, and shallow characteristics are integrated for analysis to realize the identification and the classification of characters on the image to be identified. Compared with a way that image characters are identified only through the shallow characteristics (which are obvious characteristics and can be judged by human eyes), the character identification method is characterized in that the shallow characteristic identification can be introduced to effectively increase the identification rate of the characters on the image to be identified. According to the character identification method, the deep characteristic identification and the shallow characteristic identification are combined, a robust license plate number character method is provided to improve the robustness of license plate number character identification under a monitoring environment, and license plate number identification accuracy is effectively improved.

Description

Technical field [0001] The invention relates to the field of computer technology security, in particular to a character recognition method and storage medium based on the fusion of deep features and shallow features. Background technique [0002] With the acceleration of the construction of Ping An China, the distribution of surveillance cameras has become wider and wider, and the resolution of cameras has become higher and higher. The distribution of cameras is often more random. Compared with the image acquisition device set at the standard bayonet position, the use of There are many problems with the cameras in the surveillance network (such as distributed above the road) that are distributed over the streets and lanes when performing license plates. For example, the angle of the license plate changes greatly, and the dust attached to the surface of the long-term camera will blur the collected images and increase the noise, which will lead to problems such as increased difficu...

AI Technical Summary

Problems solved by technology

For example, the angle of the license plate changes greatly, and the dust attached to the surface of the camera that has been used for a long time will make the collected image blurred and the noise will increase, which will lead to the difficulty of license plate recognition and low recognition rate.

Images