Document information extraction method and device

Abstract

The embodiment of the invention provides a document information extraction method and a device. The method comprises the steps of determining an area serving as an anchor point area in a to-be-extracted document image, adjusting an image with a to-be-extracted contour to be in the size of a preset template image, obtaining a target image, adjusting the area to be in the size of a preset template anchor point, and obtaining a target anchor point; determining the position of the target anchor point in the target image, and obtaining the position of the region of interest in the target image by combining the pre-defined region of interest of the template image with the relative position of the template anchor point; according to the size of a pre-defined region of interest in the template image, in combination with the position of the region of interest, extracting the region of interest from the target image; wherein the template image is defined according to a document image to be extracted. The embodiment of the invention is particularly suitable for extracting document information in a large number of document images with fixed formats.

Description

technical field [0001] The present invention relates to the technical field of image processing, and more specifically, to a method and device for extracting document information. Background technique [0002] In the digital age, a large number of paper documents need to be digitally archived. Document photography is a simple and effective way; but it is a problem to detect the content and quality of massive photos. [0003] Existing document identification usually includes steps such as customizing templates, type distinction, matching and positioning, area identification, and post-identification processing. First, customize the identification template for the document to be identified, including classification features and identification element information. During the identification process, the identification Feature extraction is performed on the form image, and the matching templates are screened in the template library according to the extracted features, and the best...

AI Technical Summary

Problems solved by technology

[0004] In the existing technology, when calculating the matching relationship between line segments and line segments, it is necessary to perform matching line segment by line segment, which requires a large amount of calculation and low error tolerance rate

Images