Scene text recognition method based on deep learning
In the field of text recognition,traditional OCR technology has been widely used.However,Chi-nese characters in natural scenes have the characteristics of different background shapes,distorted fonts,and complex backgrounds,which bring greater challenges to text recognition.Life is full of a large number of natural scene texts,and the application prospects are also very broad.By using the Mask TextSpotter model as the main framework for scene text recognition,and after tuning some key parameters,it has a significant effect on end-to-end and text recognition for different rules.During the implementation of the project,four phases of work were carried out,the first stage is to prepare the data and build the PyTorch environment,the second stage is to design and implement the scene text recognition algorithm based on Mask TextSpotter,the third stage is to design and implement the text recognition system,and the fourth stage is to test the system and evaluate the model.
万方数据
