Recognition System for Text Images with Uneven Illumination Based on Deep Learning

In order to improve the quality of text image recognition, this paper will combine deep learning to carry out relevant research. The research mainly aims at building a text image recognition system with uneven illumination, introduces the basic theoretical support of the system, and then launches the system design and finally tests the system. The corresponding system can be established by using the method in this paper. The system can accurately identify the text image with uneven illumination and extract the character features from it. Therefore, the system can provide help for modern information conversion and communication.