A Text Recognition Method for Local Areas of a Freight Car Based on OCR and Its Application
Abstract：To solve the problem of low recognition rate of vehicle numbers and load information in coal transport vehicles due to severely damaged printed text and distorted captured images, this paper proposes a text recognition method for local areas of a freight car based on OCR. The method detects text regions using DBNet and sorts them, then crops and corrects them using the FPGM algorithm. Image enhancement is performed using dilation algorithm and contrast adjustment, and finally, CRNN is used for text recognition. Experiments were conducted on an image dataset of 806 images. Compared with the Yolov8+OCR algorithm, the results show that this method achieves a recognition rate of 98%, an average recognition time of 3 seconds, a recall rate of 0.996, and a precision rate of 0.985, significantly improving recognition accuracy and efficiency.
Machine-generated Keywords：
Publication Date：2025-09-30
Online Publishing Date：2025-10-21(First online date of this platform, not the publication date of the document)
Pages：5( 79-83 )