Overview / Usage

Increased levels of cholesterol have become a major concern nowadays. In this project, we tried to develop a framework for efficiently indicating if a person has high cholesterol levels by using their Android mobile phones. The edges of the iris start developing a greyish color when there are high cholesterol levels in an individual. Based on this phenomenon, Image processing techniques are used to analyze and determine the levels of cholesterol. The user can upload the photo directly by clicking or from the gallery.
There are other features also available at the Android end, namely, a planner for keeping track of medical appointments, an FAQ and tips section, a google maps API integrated framework for showing the nearest Gastrointestinal clinics or hospitals.

Methodology / Approach

IMPLEMENTED:
The approach used for determining cholesterol levels is as follows,

1. The image is passed through Gaussian and Bilateral filters to reduce the effect of noise.
2. The pupil is detected by thresholding techniques. (As pupil gas a deep black color always.)
3. Iris is detected by predicting the radius of the iris based on the radius of the pupil.
4. The useless part of the image is cropped out and masks are applied to obtain the useful pixels in the remaining image.

The approach used for developing the app is as follows,

1. The user can upload his/ her image onto the server using Intent and sharedPreferences.(Action_pick)

PROPOSED:

1. FAQ and Tips
2. Google Maps API for showing markers of nearby clinics.
3. Alert system for sending Appointment Notifications.

Technologies Used

1. Python (OpenCV, Sklearn, Scipy, Numpy, Pyrebase, Sys, Os )
2. Android Studio (Java)
3. Firebase
4. Google APIs

Repository

https://github.com/adi-797/Hacktillend-BVCoE

Other links

• Java Files.
• Publication