Eye Gaze

Project Title

Eye Gaze

Project Area of Specialization Artificial IntelligenceProject Summary

Eyes provides the greatest support to the disabled so we are providing facility or helping those victims by the use of their own eyes.

Disabled people are unable to do any type of work. Disabled people are not able to control any type of device due to their disability. How would they get employed? they can, because of a very advance technology through eye gaze movement detection system.

By using Bluetooth we have connected our eye gaze module with Remote Control Car and the accuracy of 5 points which are left right top down and centre are very smooth.

in future we can use eye detection module to implement different types of application, we will use eye detection module to provide security, to control IOT base solution.

also we can make quad copter and we can move it with our eyes

Project Objectives

We are providing a technical way to achieve our aim by using our Eye Detection System. This will provide an effective way to interact with computer and which is made for disabled one's.

 Objectives

1-Eye Detection System is also providing a new way of controlling devices like self-made cars, toys which results in advancement of Arduino Projects

2-To make an Eye Detection System with naked eye.

3-Detection of Five Movement of Pupil which include “Up”, “Down”, “Left”, “Right” and “Center”

Project Implementation Method

IMPLEMENTATION

Threshold Function:

This function is named as threshold because it means that the image we are getting in grey scale it will convert it into black and white image.

We will pass this image to threshold function and then this function will return two arrays that is point x and point y than we will pass these two arrays to another function which will track the eye position. This threshold function is used to create a line between eye to point out the coordinates of Eye.

Usage of cut-off variable:

When image is converted into binary, we get some black pixels which are not useful so we will ignore these pixels (approx. 5 pixels) because if we will not ignore them then it will track whole eye and it is obvious that eye will never move to the most corner position while moving. This variable is used to get efficient tracking of eye.

Position Finding:

We are detecting 8 Position of Pupil, which include:

• Upward:
  • Algorithm first detect the centre position of pupil, when pupil start moving upward it compare its current position with the centre point, Y-Axis value decrease according to Python Image Processing.

• Left:
  • Algorithm first detect the centre position of pupil, when pupil start moving toward left it compare its current position with the centre point, X-Axis value decreases according to Python Image Processing.
• Right:
  •  Algorithm first detect the centre position of pupil, when pupil start moving toward left it compare its current position with the centre point, X-Axis value increases according to Python Image Processing.
• Downward:
  • In Downward detection, when we see downward our eye lid start closing so the detection of a pupil will be hard to find, so we need to get another way to detect downward positon in efficient way, for that we use the function called “EYE ASPECT RATIO” which detect the closing of eye lid which we consider as downward position.

Eye Aspect Ratio:

Formula used for eye aspect ratio:

Benefits of the Project

Handicap people can get benefit from this eye detection module

Disabled IT Professional can work and interact with different technologies using this eye detection module .Not only this we can implement this eye module for different implementations.

In future, eye tracking need to develop standard for matrices used for detecting eye gaze, which should have how they will be used in UI designs. For Example, unwanted equipment should be removed to make user more comfortable and accuracy should be increased with higher resolution of camera. Process of tracking the eye should be cheaper so companies can implement that in their products and can be easily commercialized 

Technical Details of Final Deliverable

The objective of this project is to develop Eye detection system to recognize the movement of Eye Pupil. This report explores unique techniques used for the recognition of Pupil. Different stages involving image processing like the pre-processing stage, face recognition and feature extraction will be studied and discussed. Finally, this Eye detection system will be implemented and written in the language called "Python”.

This project uses the Python Language to develop the software. The main advantage of using this Language is that it provides easy way to detect image that is suitable for character recognition. This software uses the world’s simplest face recognition library d-lib and Open-Cv. Using them we proposed an Eye Detection Method which allows to track and detect eight different positions of Eye pupil that is left, right, up, down, left up, left down, right up, right down.

Final Deliverable of the Project Hardware SystemType of Industry IT Technologies Artificial Intelligence(AI), RoboticsSustainable Development Goals Sustainable Cities and CommunitiesRequired Resources

Item Name	Type	No. of Units	Per Unit Cost (in Rs)	Total (in Rs)
			Total in (Rs)	60000
Rc Controller Car	Equipment	1	20000	20000
Quad Copter	Equipment	1	40000	40000