Container Number Recognition System Based on Improved DBNet and Improved CRNN
Aiming at the inaccurate positioning of the current container number recognition algorithm and the weak ability to recognize skewed and distorted text,a container number recognition algorithm based on improved DBNet and improved CRNN is proposed in this paper.The attention mechanism is introduced in the feature extraction network of DBNet,which effectively improves its text positioning ability.The spatial transformation network is introduced in CRNN,which enhances its ability to recognize skewed and distorted text.The combination of text location and recognition model in series reasoning has achieved a recognition rate of 98.3％in the test scenario,which is of practical value.
万方数据
维普
