Table type determination method and device based on neural network model, and medium

Abstract

The invention relates to the technical field of artificial intelligence, and provides a table type judgment method and system based on a neural network model, equipment and a storage medium. Comprising the following steps: inputting first form information into a correction network to obtain first correction result information, and inputting the first correction result information into a framework feature extraction network to obtain first form line feature information and first cell feature information; constructing a first positioning coordinate system by the first table line feature information, inputting the first table cell feature information to generate first table cell positioning information, traversing a first table cell set to carry out row and column vector identification, obtaining a first table cell row vector and a first table cell column vector, and inputting the first table cell row vector and the first table cell column vector into a row and column feature extraction network; and inputting the first semantic feature information and the first layout feature information into a table type identification network to carry out table type optimization screening, and obtaining a first screening result to generate a first table type judgment result. The invention further relates to the technical field of block chains, and the table type can be stored in a node of a block chain.

Description

technical field [0001] The present application relates to the technical field of artificial intelligence, and in particular, to a method, system, device and storage medium for determining a table type based on a neural network model. Background technique [0002] In many industries, such as the financial industry and the engineering industry, a large number of forms need to be faced. In the past, manual processing and formulation of forms was required. With the rapid development of artificial intelligence, due to the regularity of the form itself, it has suffered a lot. Note that among these, the identification of table types is the basis for table data processing, and the type determination of table layouts plays an important role in subsequent understanding and extraction logic. [0003] The common table type determination in the prior art is mostly based on rule-based systems, or uses shallow classifiers based on artificially designed feature training to identify table ty...

AI Technical Summary

Problems solved by technology

[0003] The common form type determination in the prior art is mostly based on the rule system, or using a shallow classifier trained based on artificially designed features for form type identification, but because the manual setting of classification rules depends on expert experience, there is migration and accuracy Difficult to guarantee

Images