Team Name: Developers Inc.
Team Leader: Abhinav Singh
Team Members:
Got to know about Hive through the Hacknovate5 event. The fun activities were interesting and students learnt about blockchain and got some freebies.
Problem Statement: Inadequate ID verification of a person whether he/she is wearing the ID card or not.
From Where we Got this Idea?: When a student reaches their respective institution, they have to scan their face or ID or they have to show their ID card to the Guard which takes a lot of time of the students (you know engineers they're already late so😛). To overcome this problem and make it hassle free, we came up with the solution of this Problem with a ML ID Analyzer which is called "ScanSage".
Problems We Ran Into: One of the biggest challenges was to find the best possible dataset for training the model on humans and id cards, finding a right database that could hold information to our likings and such that we landed on MongoDB.
Finding the correct set of conditions which are able to detect Ids on people somewhat accurately in such a short amount of time was also a challenge. We also gave a lot of thought to the real world applications of this software and would it be a somewhat feasible business model.
Solution: The project aims on detecting the ID Card of a student through a ML algorithm YOLOv8. An Android Application ScanSage is used to keep track and of the real time data of the person whether he/she is wearing the ID card or not. The data is stored and retrieved through MongoDB which implements the CRUD operations. It enhances the security.
ScanSage comprises two main components:
ML Model Component: This component is responsible for detecting whether a person captured by a CCTV or webcam is wearing an ID card or not. Upon detection, the relevant data is uploaded to a MongoDB database.
Android Application Component: The Android application syncs data from the MongoDB database in real-time using a WebSocket instance. It provides real-time analytics through graphs and an intuitive user interface. The application includes features such as authentication through email and password, searching entries of persons wearing ID cards or not, and generating notifications whenever a person is detected without an ID card.
Technologies Used: Yolo v8, Python, Open CV, Pandas, Matplotlib, PyTorch, MongoDB Realm, MongoDB Compass, Jetpack Compose, Kotlin, Kotlin Coroutines
Project Screenshots:
App Screenshots:
Model Screenshots:
Event Picture:
Demo Video:
Event Experience: The experience was amazing as we got to learn about various technologies along with team & time management skills. The fun activities rejuvenated the students and freshen our minds. The competitive environment kept the determination & dedication among us to complete our project within the deadline. Me & my team will participate in the next Hacknovate event with gained experience to implement it perfectly.😁
Links:
