基于PaddleOCR识别线缆物资标签的研究

打开文本图片集

中图分类号：TP391.4 文献标识码：A 文章编号：2096-4706（2025）09-0128-04

Abstract： With the rapid economic and social development in China， more and more cable materials are produced and preserved for construction， but the large number of cable materials poses a serious challenge to the statistics and management of managers. This paper researches an efficient cable material label recognition model to solve the problems of inefficiency and error-prone nature in the cable management process. Image preprocessing is carried out by using Grayscale Image Processing， Binarized Image， Morphological Processing， Connectivity Domain Labeling， Edge Detection and other image processing techniques， and then OCR is used to extract text information so as to save effective information. The experimental results show that the proposed method performs well in label recognition， and the accuracy of label text information recognition reaches 95% . This research provides practical and technical support for the automated management of cable materials， and also provides an important reference for further research in related fields.

Keywords： image processing; PaddleOCR; automated management

0 引 言

线缆物资标签主要用于介绍线缆的基本信息、产品参数和生产许可等相关情况。（剩余6061字）