Core demands
Government approval is often a synonym for tedious and troublesome. On the one hand, various paper information needs to be entered into the system one by one in the approval process, and on the other hand, there are high standards for timeliness and accuracy. Therefore, the input workload of the electronic license information to be approved is very large, and there are risks in the speed and accuracy of long-term dependence on manpower. At the same time, since most of the materials submitted are personal certificates and enterprise certificates, information confidentiality is also an important factor for the political review unit to consider. Therefore, how to effectively convert a wide range of paper certificates into electronic certificates while ensuring the confidentiality of information has become an urgent problem.
Solution
The electronic license approval system developed by Huizhengtong can realize the user-defined configuration of the license template and automatically extract the image and text information on the certificate after it is connected to the character recognition of Baidu Brain iOCR user-defined template. After the electronic license is generated, the subsequent signature business can be easily handled by manual review/modification. The specific implementation process is as follows:
Deployment process: Baidu Brain provides iOCR model capability; Huizhengtong deployed and ran the privatization deployment package and developed it into an electronic license processing system for the second time. Business process: the government processing party directly uses the system for e-government approval. The iOCR self defined template text recognition capability can freely create recognition templates for files with different formats according to business needs. IOCR template management interface:
IOCR user-defined template text recognition supports the selection of user-defined recognition fields; The fields to be identified can be selected according to the identification requirement box. IOCR template creation interface:
Use the created template to identify the text content of the corresponding file, as shown in the figure: