Method and system for extracting key information from identity card image picture

Abstract

The invention provides a method and system for extracting key information from an identity card image picture, and the method comprises the steps: obtaining an identity card picture through recognition, detecting the identity card picture through a face detection model and a national emblem detection model to obtain front and back screenshots of an identity card, carrying out the textbox recognition of the screenshots with shading removed, obtaining a textbox screenshot, and performing character recognition on the textbox screenshot, and calibrating a recognition result to obtain a final recognition result. By adopting a multi-model series processing structure, the problem of identity card picture information extraction under various picture quality conditions is solved; the problem that skewed and inverted images cannot be recognized is solved through rotation recognition. The identity card number training data annotation problem is solved by utilizing the verification rule of the identity card number and combining with a plurality of machine learning models for common identification. By using the image morphology to cooperate with the CTPN model, the problem of CTPN model training set data annotation in the textbox recognition process is solved.

Description

technical field [0001] The present invention relates to the technical field of image processing, in particular to a method and system for extracting key information from ID card images. Background technique [0002] Patent document CN109145891A discloses a client and its method for identifying an ID card, and a system for identifying an ID card. The method for the client to identify the ID card includes: receiving the image of the ID card in real time; identifying the first area according to the first preset set; the first preset set is stored in the client and includes picture, the first area is the area at the preset position of the ID card image; if the ID card number is successfully identified, the area to be identified in the ID card image is determined according to the preset position; the information related to the area to be identified is sent to The server is configured to identify the content in the area to be identified according to the received information. [...

AI Technical Summary

Problems solved by technology

[0004] 1. The existing ID card picture information extraction methods mainly focus on image shooting and recognition of identity proof objects, which have high requirements for picture clarity, and cannot solve the recognition problem when the picture is rotated and inverted, and the applicable scenarios are limited.

[0005] 2. At present, the traditional technology of image text recognition using machine learning needs to label a large number of training samples, and most of the work needs to be done manually.

[0006] 3. The existing ID card picture information extraction method lacks the information correction function, and fails to further correct all kinds of information on the extracted ID card picture to improve the accuracy of recognition

Images