Method and device for establishing bill type identification model and method and device for identifying bill type

Abstract

The invention discloses a method for establishing a bill type identification model and a method for identifying a bill type, relates to the field of artificial intelligence, in particular to a computer vision and deep learning technology, and can be used in an OCR (Optical Character Recognition) scene. The method for establishing the identification model comprises the steps of obtaining training data; performing text detection on the plurality of bill images, and determining a textbox in each bill image and position information and text information of each textbox; constructing a neural network model containing a multi-modal feature extraction module; and training the neural network model by using each bill image, the position information and the text information of each textbox in each bill image, and the labeling type to obtain an identification model. The bill type identification method comprises the steps of obtaining a to-be-identified bill image; performing text detection on the to-be-recognized bill image to determine textboxes and position information and text information of each textbox; taking the to-be-recognized bill image and the position information and the text information of each textbox as input of a recognition model, and taking an output result as a bill type.

Description

technical field [0001] This application relates to the field of artificial intelligence technology, specifically computer vision and deep learning technology, which can be used in OCR scenarios. Provided are a method, device, electronic equipment and readable storage medium for establishing a bill type identification model and identifying the bill type. Background technique [0002] Bill is a text carrier of important structured information. With the development of social forms and the increase of commercial activities, the styles of bills have shown a linear increase, and various types have been developed. When the relevant departments conduct reimbursement, several or even more than a dozen different types of bills need to be reviewed, and the structures of some bills are very similar. Relevant personnel need to finely classify these bills according to the audit requirements, which is time-consuming and labor-intensive. Contents of the invention [0003] The present a...

AI Technical Summary

Problems solved by technology

Relevant personnel need to finely classify these bills according to the audit requirements, which is time-consuming and laborious

Images