Financial bill seal erasing method based on attention mechanism and generative adversarial network

Abstract

The invention belongs to the field of bill text recognition, and relates to a financial bill seal erasing method based on an attention mechanism and a generative adversarial network, which is implemented by a processor and comprises the following steps: receiving an original picture of a financial bill; determining a first feature map according to the original picture by using a feature extraction module in a convolutional neural network; using the convolutional neural network to respectively extract a background color map of the original picture and an attention heat map reflecting position distribution of a stamp on the original picture according to the first feature map; using the convolutional neural network to generate a seal-erased picture of the original picture in a generative adversarial mode according to a second feature map spliced by the original picture, the background color map and the attention heat map in the channel direction; training the convolutional neural network in a generative adversarial mode. The problem that the financial bill containing the seal is difficult to identify is solved, and the purpose that the seal is erased without losing original character information is achieved.

Description

technical field [0001] The invention belongs to the technical field of graph convolutional neural networks, and in particular relates to a method for erasing and filling partial areas from a picture. Background technique [0002] In the process of automatic computer processing of financial bill reimbursement, the starting point of the processing process involves the digital entry of financial bills. Physical financial bills include invoices, train tickets, air tickets, and approval forms. Various types of business documents. The method of digital entry of financial bill reimbursement is to scan financial bills into digital image files through image capture devices such as scanners, and then use algorithm models to detect and identify the content information of financial bills from digital image files. A practical problem is that the original physical financial bills are basically stamped with seals. These seals randomly cover, split, and mix content information, resulting in...

AI Technical Summary

Problems solved by technology

[0004] The purpose of the present invention is to address the defects of the current methods in the background technology, and provide a method based on the attention mechanism in deep learning and the generation of confrontation network to erase the seals in the financial bills, and solve the problem specially through the attention mechanism and the generation of confrontation network Difficult identification of financial bills containing seals, achieving the goal of erasing financial bill seals without losing original text information, promoting financial office informatization by solving the problem of financial bill seal erasure, saving social human resource costs and simplifying reimbursement processing procedures

Images