Identification method based on classifiers in image document electronic material identification system

Abstract

The invention provides an identification method based on classifiers in an image document electronic material identification system. The classifiers are arranged in the identification system, identification information of images is classified to obtain different information items, corresponding lookup tables are built for all information items, and the identification information is compared through content in the lookup tables. The method can automatically identify the scanning images, useful information is extracted from the scanning images and is stored in a database according to a certain classifying rule for user searching and inquiring, and workloads of users are reduced to the maximum extent. The rate of identification of characters is improved through a multi-classifier fusion method; a format template is adopted, the content of the different information items are compared through a multi-region and multi-content redundancy check method, the sufficient reliability of an identification result is ensured, and the identification efficiency is improved.

Description

technical field [0001] The invention relates to the field of data management systems, in particular to a method for identifying classifiers in an electronic data identification system based on image files. Background technique [0002] In modern society, paper documents (such as bank vouchers, personal information forms, etc.) are still widely used, and it is very difficult to store and manage paper documents, and to classify and find information on documents. The popularity of computers and smart phones has made it possible to manage paper documents electronically, but it takes a lot of time and manpower to manually input the information on paper documents into the electronic system; and the automatic identification of bill content through intelligent systems is still There are many limitations. [0003] For example, in banking business, a large amount of information on the bills are printed numbers and Chinese and English characters. Accurate extraction and identification...

AI Technical Summary

Problems solved by technology

However, due to the complexity of the bill layout and the particularity of the recognition requirements, various difficulties may be encountered in the actual system: there are interference information such as stamps, inks, handwritten information, background patterns, etc. on the bill layout; There are problems such as character sticking, frequent font size changes, and incomplete identification information

Images