微型电脑应用2025，Vol.41Issue(9)：174-177,4.
基于DSCC算法和字符特征的气象纸质表格快速数字化方法
Fast Digitization Method for Meteorological Paper Tables Based on DSCC Algorithm and Character Features
朱津辉 1汤津赢 1王帅 1单薇薇 2冯冬蕾3
作者信息
- 1. 辽宁省气象信息中心,辽宁,沈阳 110166
- 2. 辽宁省气象装备保障中心,辽宁,沈阳 110166||辽宁省气象信息中心,辽宁,沈阳 110166
- 3. 辽宁省气象台,辽宁,沈阳 110166
- 折叠
摘要
Abstract
The paper meteorological tables has problems,such as non-standard writing handwriting pressing and alteration re-sulting in slow and low accuracy in digital conversion.To address the above problems,a fast digitization method for meteoro-logical paper tables based on directional single connected chain(DSCC)algorithm and character features is studied.A scanner is used to collect meteorological paper table images and implement preprocessing,DSCC algorithm is used to eliminate table line segments and separate text lines from the background,the smallest univalue segment assimilating nucleus(SUSAN)algorithm is used to extract character corner features and determine the position of the character area through the minimum bounding rec-tangle.The Le Net-5 structure is used to recognize characters and convert the characters in paper tables into electronic format data on the computer,thus completing the rapid digitization of meteorological paper tables.The results show that using the proposed method,the character misrecognition rate and misrecognition rate are the lowest and the time spent is the least.Therefore,the proposed method can complete the rapid digital conversion of meteorological paper tables more quickly and accu-rately.关键词
DSCC算法/字符特征/气象纸质表格/字符区域定位/数字化方法Key words
DSCC algorithm/character feature/meteorological paper table/character region localization/digital method分类
计算机与自动化引用本文复制引用
朱津辉,汤津赢,王帅,单薇薇,冯冬蕾..基于DSCC算法和字符特征的气象纸质表格快速数字化方法[J].微型电脑应用,2025,41(9):174-177,4.
