CASE STUDIES

ARTIFICIAL INTELLIGENCE / MACHINE LEARNING

Indian Currency Note Detection Using Deep Learning (CNN) and OpenCV with Python

Python 3.7, Keras, OpenCV

The objective of this project is to identify various Indian currency notes using Convolution Neural Network(CNN) which are placed in front of a camera. It’s obvious application of for enabling Blind people to handle financial transaction using their mobile phone (where mobile camera will take a live feed of currency and Python ML will identify the currency value ) and also helping world traveller to easily identify currency notes from different country (which is difficult to separate out in case of certain countries)

In this project we train the datasets to identify 10, 20, 200, 500 Rs Indian currency notes with very high accuracy. We can use both webcam or images from the camera to predict the value of the notes.

The further extension of this project would be to add more notes of Indian currency as well as for the other countries (eg USA, UAE, etc.)

Car Driver Drowsiness Detection using Computer Vision (Using Facial Mapping)

Python 3.7, OpenCV, Scipy

Road Accidents are one of the top reasons for death worldwide. According to WHO Road accidents killed 1.4 million people in 2016. One of the major reasons for road accidents can be because of drowsiness. According to Central Road Research Institute (CRRI) India, 40% of highway accidents occur due to drivers dozing off.

(https://www.financialexpress.com/india-news/40-of-highway-accidents-occur-due-to-drivers-dozing-off/1659901/)

The best way to prevent this is to install a webcam at the dashboard of the car, which can automatically determine if a person's eyes are closed for more than 3 seconds(that means that driver has dozed off) and blow off an alarm, to alert the driver.

This project serves as a solution to this problem using python libraries.

We have used OpenCV and other similar libraries to detect the position of the eye and continuously check if the eyes are closed for a certain amount of time.