Automated Identity Document Recognition and Classification (AIDRAC)-A Review

In this technological era, official documents like Government approved IDs, Certificates, and other documents are paper-based or image based. The problem arises when an organization has to go through each of these documents manually and extract, enter or search for information. Automated Identity Document Recognition and Classification (AIDRAC) is one of the techniques to tackle this problem by decreasing manual effort. This kind of model automatically classifies government-issued identity documents into predefined categories such as AADHAR card, PAN card, and Driving License. It also recognizes and extracts details from the documents uploaded, providing the users the option to autofill their details upon scanning. This can be achieved through the application of Image Processing algorithms using Machine Learning. The extracted information can be stored succeeding relevant access and permissions. The data can be viewed by authenticated consumers when required. Due to the automation of manual work, the implementation of this model saves time, thus saving resources. This paper presents an overview of the exhaustive survey of the techniques employed to build such models and, a comparative analysis of these techniques. It has been found from the survey that there are no models existing that completely accomplish the said automation. Our future research is to build an automation model that can be used in various organizations like Banks, Education Institutions, Law firms, Travel Agencies etc. to autofill the details extracted from the scanned images of Identity proofs. The Sequence diagram of the proposed system is also presented in this paper. The proposed system is expected to outperform the existing models.