-
-
A Secure Document Verifier
Inspiration
The Internet has become a sensation nowadays no doubt we are growing faster in the world of computer and internet, as each coin have two sides so does the internet. We all are aware of the good side but another side the cybercrimes are also committed through the internet. And many images processing and deep learning tools are accessible to produce digital evidence in criminal activities. There are many fields where technology is used as a weapon for committing cyber crimes like fake currency notes, cheques, as well as manipulation of important government documents, like Aadhar Card, Driving Licence and Signature or educational certificates. To verify and check them individually is not an easy task but we can utilize the technology of image processing and deep learning model to overcome this process
What it does
So we come up with our proposal to provide an efficient and accurate image processing and deep learning model inside a simple and attractive UI based application with a scalable and durable database system that is capable of storing a lot of data into it which recognises the counterfeiting and computer manipulated fraud and provide the predicted result which will help in the collection of evidence in cybercrime, where user or industry can simply register and use our platform in just one click.
How we built it
Deep learning: we use deep learning models for the verification of the documents and also many packages for the processing of images so that they can be matched with the stored images or documents. The packages are NumPy for converting the documents in the array, sklearn for train test and also for finding accuracy of the model. TensorFlow for training the model (many computations are done using TensorFlow), OpenCV for image processing and many other libraries likes Keras And various tools for training the model with a proper modulation or accuracy.
Chatbot: NLP based chatbot to answer all query, related to document verifications and platform issue and help you 24x7.
Back-End: We are using python framework Django for handling the images and connecting the deep learning model through the frontend website and the high scalable database like SQLite Or PostgreSQL for storing the data and documentation of the user
Front-End:: We are using a JavaScript framework React js and CSS with SCSS and material design with different dynamic UI themes for the attractive webpage.
Challenges we ran into
Our application depends on the person's image, signature, Aadhar number, pan number, in the case of a new document it will depend on document ID and user name.
Accomplishments that we're proud of
We Participated in the Trithon 2021 hackathon organized by the Trident Academy of Technology Bhubaneswar where more than 80+ teams applied and 26 teams selected and worked on problem statements given by the college and developed their MVP in 24 hours we also participated and developed our application using AI ML along with Natural language processing to verify the documents like Aadhar card pen and mark sheet in this short span of time we developed our whole application using backends like Django and SQLite database.
1st Round: It was about Introduction about team statement, solution and solution validation along with the need
2nd Round: This round was all about our progress and teamwork we have shown our train models and algorithm implementation apart from this we assigned the task of integrating all in an application.
3rd Round: This was the last round of evaluation where we present the whole application and working process with different features
A lot of positive feedbacks from professional people working in TCS and TAT
What we learned
It's just an incredible feeling that we have been a part of this coding event. We have spent many endless nights being awake to achieve this wonderful position by all our teammates !! we learnt a lot of things apart from coding such as team-playing and pitching business management
What's next for dAuth
This Problem statement was provided by DRDO so the team is currently trying to pitch the proposal to them and the hand we are working on making a business model for starting the startup
Log in or sign up for Devpost to join the conversation.