A key text detection and classification training method for certificate pictures

Abstract

The invention discloses a method for key text detection and classification training of certificate pictures. The method includes the following steps: S1, constructing a certificate template, and generating training samples based on the certificate template; S2, constructing an algorithm model integrating text detection and classification, and Output the text area and its classification results based on the detection feature map and classification feature map output by the network; S3, import the training samples into the integrated algorithm model of text detection and classification for training, and respectively analyze the classification feature map and detection prediction feature map during the training process Carry out two types of negative sampling; S4, sequentially calculate the classification feature map loss value in the classification feature map and the detection prediction feature map loss value in the detection prediction feature map, and fuse the classification feature map loss value and the detection prediction feature map loss value to reverse Optimize the training text detection and classification integrated algorithm model to the propagation. Beneficial effect: the training of deep neural network can be supported by collecting a small number of samples.

Description

technical field [0001] The invention relates to the technical field of text detection, in particular to a key text detection and classification training method for certificate pictures. Background technique [0002] In the text recognition scene of certificate pictures, it is necessary to extract the text content of the area of ​​interest from the picture and classify it into the corresponding fields. There are two common methods: 1. Combining general text detection and general text recognition to obtain the text content in the picture After the position and content of all texts, the recognition results are divided into the corresponding content fields based on the content rules of different fields; 2. Extract the region of interest based on the template reference position, first calibrate multiple reference positions on the template, and in actual operation First find out the reference position in the picture, then perform text detection, and determine which field each text...

AI Technical Summary

Problems solved by technology

[0003] The main problem of the former is that it involves the writing of a large number of text rules in the field division stage, the text content of some fields has no obvious rules, and there is a lot of noise in the actual scene, which makes the overall key text detection and classification results unreliable; The dependence of the field is large. When there is a lack of reference position or deformation of the picture, the field classification result is unreliable.

Images