Point of Assessment - B094

using machine learning to detected help seeker by using algorithm to scan security cameras searching for arm waving

Comment

Inspiration

  • simple , effective solution with no new hardware installation required.
  • easily used by all ages
  • cross language barrier

What it does

using machine learning to detect and locate help seeker by using machine learning algorithm to scan security cameras searching for arm waving as sign used by Pilgrim to ask for help.

Pre-Hajj procedures:

  • Inform the Pilgrim how to ask for help by direct them self to a camera and wave arms

Workflow of the system

1- machine learning algorithm scan security and TV cameras visual feed to detect arm waving. 2- help seeker detected and accurate location is sent to nearby general help provider nearby according to ( GPS location of the help seeker ).

3- [Mobile APP] General help seeker send feedback contain the following:

  • categorize the help needed ( main category as: medical assessment , transportation, Lost ,simple help - something can be done by the general help seeker and need no farther assessment ).
  • urgency of the help ( red , yellow , green)
  • new location or confirm location ( by GPS )
  • picture of help seeker.
  • case title
  • notes ( voice note or written notes ).

4- [Online system ] each help category will have online dashboard that provide the following :

  • charts about requests
  • map show requests location and specialized help provider location.
  • view requests details
  • follow up requests .
  • assign requests to specialized help provider

5- [ Mobile App ] notification is sent when a new request is assigned to specialized help provider with details and location .

6- feedback is updated by specialized help provider

How I built it

  • accurate GPS location of each camera most provided .
  • Area dimensions and angels that each camera cover.
  • create data sets for machine learning algorithm.
  • make machine learning algorithm.
  • create mobile app .
  • create online system.

Challenges I ran into

  • develop efficient algorithm to detect the help seeker
  • false detection
  • low light situation
  • cloth cover the arms.
  • provide accurate location

What's next for Point of Assessment

working on including other arm movments to indicate different kind of help.

  • include night vision visual feed to improve low light situation condition
  • use drones that has speaker , mic and camera to achieve the general assessment step.
Share this project:

Updates