Template-based key value pair extraction method and system

• Main Authors: LU CHENG, CHEN WUCONG, CHEN JINFENG, ZHU JIAHAO, DENG YAOLONG, ZHONG JIANCHEN, LIANG PEISEN, LI WENHAO
• Format: Patent
• Language: Chinese; English
• Published: 31.12.2021
• Subjects: CALCULATING; COMPUTING; COUNTING; HANDLING RECORD CARRIERS; IMAGE DATA PROCESSING OR GENERATION, IN GENERAL; PHYSICS; PRESENTATION OF DATA; RECOGNITION OF DATA; RECORD CARRIERS

The invention discloses a template-based key value pair extraction method and system, and the method comprises the steps: receiving an image, finding text information in the image through a text detection algorithm DB, and forming a bounding box around the text information; finding key word information in the image according to a pre-input template, and calculating a fixed anchor point according to the key word information and the coordinates of the bounding box; calculating variable anchor points according to the coordinates of the bounding box and the fixed anchor points; performing projection transformation correction on the image according to the fixed anchor point, the variable anchor point and the template; and identifying and extracting the text information in the corrected bounding box. According to a keyword or a regular expression or a user-defined anchor function, projection transformation correction is carried out according to anchor information, the method is suitable for correction and extractio