Image processing apparatus, image processing method, program, and storage medium

摘要

Each page in document images is divided into regions. Attribute information is assigned to each divided region. Character recognition is performed for a region having a caption attribute, thereby obtaining character information. Metadata of an object is extracted from a region containing an expression synonymous with the character information. Even when the object and metadata exist in different pages, the metadata is associated with the object. The document image is transmitted for each page. Finally, the metadata is transmitted.