Project Proposal - Instructions

 

Project proposal should be about 1 page long

 

1) Title

2) Brief Description of Project goals - what will the final system do.

3) Methods.

    a) Hardware you will be using

    b) Describe the various system parts, what system/codes/toolboxes you will rely on and integrate
         into your system.

    c) Describe how these modules work

    d) Expected Challenges

    e) Possible solutions for the challenges

    f) Other methods that were considered

4) Time line plan

 

 


 

Example Proposal

 

Title: Face Recognition System

End Goal:

A real time system to recognize faces.

Two Operation Modes:

1) Training:

    a) given a set of images + associated information (name etc), adds them to DB

    b) A single user interactively viewed through system camera is captured and image is added to DB.

        User also supplies associated information.

2) Run time:

    Users face the system camera and the system attempts to recognize the

    user. System provides Audio + Text response on screen.

    If user is not recognized system requests user to initiate image acquisition for new user.

 

Methods:

RGB web cam and laptop will be used.

The system will rely on OpenPose Face detection functions, together with OpenCV Face training and learning.

DB management will be based on Pythons DBtool.

Project will be implemented in Python

 

How they work:

Face detection is based on Viola Jones face recognition.

Training and retrieval use Bag of Words.

 

Challenges:

The system must work under various illumination conditions.

System must allow large variations in appearance including:

face expression, hair cut, glasses, hats and more.

 

Possible solutions to Challenges:

Normalize input images, (color and lighting) and perform image enhancement.

Allow multi classes for a single person.

Possibly will require modules for eye glass detection

 

Other methods that were considered :

OpenCV's raw face recognition functions.

Python's Face recognition API

Matlab's Face recognition system.

 

Time line Plan:

* Camera installation and setup - 1 week

* Code installation - 1 week

* Testing on sample images - 2 weeks

* Mid proposal demo

* Real world testing - will place system in lab - have students test the system - 2 weeks

* Rap up - Rap up the system for export:  - 1 week

      * Clean code, add comments and documentation

     *  Write up installation instructions

      *Write up user instructions

      * Write up Project report