基于OCR的天地图·湖北行政村(居)委会质检关键技术研究

Research on Quality Inspection Key Technology of“MapWorld·Hubei”Administrative Village(Neighborhood)Committee Based on OCR

针对湖北省地理信息公共服务平台(简称“天地图·湖北”)行政村(居)委会地名地址常规人工质检效率不高、智能化应用缺失等问题,提出一种基于OCR技术的智能地名地址识别方法,并针对实施过程中的图像处理、文字识别等环节中的关键技术进行了深入研究。通过开源深度学习模型完成的工程实例表明,智能地址识别方法在应用效率上有显著提升,该方法的推广使用在“天地图·湖北”地名地址质检工作中具备较大的应用价值。

In view of the low efficiency of conventional manual quality inspection and the lack of intelligent application of place name and address of administrative village(neighborhood)committee of Hubei geographical information public service platform(“MapWorld·Hubei”),we proposed an intelligent place name and address recognition method based on OCR technology,and deeply studied the key technologies in image processing,character recognition and other links in the implementation process.The engineering example completed by the open source deep learning model shows that this method has significantly improved the application efficiency,and the popularization and application of the method has great application value in the quality inspection of“MapWorld·Hubei”place name and address.