Docstoc

Human Detection and Activity Monitoring - Waqas Burney

Document Sample
Human Detection and Activity Monitoring - Waqas Burney Powered By Docstoc
					codename: imaGInation KIeve
FYP Presentation – the Finale!
 24.04.08

Project Team :
 Syed Waqas A Burney, 2004185, FCSE (Team Lead)
 Mutahira Ikram Khan, 2004136, FCSE


Project Advisor :
 Mr. Badre Munir, FCSE
Agenda :

 Core Progeny
 Project Objective
 Project Modularization
 The System
         Overview
         Architecture
         Module 1 – Motion Detection & Human Activity Monitoring
         Module 2 – The Decision Making Software Engine
         Module 3 – Appliance Control Mechanism & Prototype Modeling
 Project Status
 Envisioning Ahead


                                                            vSmart Demo
Core Progeny

 Computer Vision
    The branch of applied computing that concerns modeling and developing
     artificial systems that attain information from images and thereby provide
     useful functionality and results.
 Domotics
   Study of specific automation requirements for homes and application of
     automation techniques for the comfort, convenience and security of the
     residents



Development Platform :
        .NET 2.0 Framework (C#)    Visual Studio 2005   MS SQL Server 2005
Objective

  Prototyping an integrated system for the smart control of domestic
  electronic appliances in local environments, based on monitored user
  presence and activity; thereby achieving multidimensional user-
  convenience and using modern technology to endow a sustainable
  environment in our everyday lives.


Applications :
 User Convenience and Energy Saving
    Offices (lunch hours, after-office timings, etc)
    Houses (the ‘Digital Home’ concept)
Modularization (and timeline)
 Problem Statement




                                                                                        Jun
 Ideation through innovative techniques/lateral thinking/ brain storming
 Engineering solutions, and thereby outlining the very FYP definition, field & scope




                                                                                              codename: imaGInation KIeve
 Adopting a research based/experimental approach




                                                                                        Nov
 System design




                                                                                        Dec
 Motion detection and respective human activity monitoring
 Prototype modeling




                                                                                        Jan
 Serially interfaced appliance control mechanism




                                                                                        Mar
 Decision taking software engine– a multi-threaded expert system




                                                                                        Apr
 Optimization and deployment
 Completion
System Overview
                                                                    Local Environment
                                                 Module 1
                                      Human Activity Monitoring (Input)

                                                          CMOS Webcams

                                   User
                                  Activity




                      Software
                       Engine
                                              Domestic
                                             Appliances
     Module 2
The Software Engine
   (Processing)
                                         Module 3
                                 Appliance Control (Output)
System Architecture
System Architecture




                      MODULE 1
  MODULE 1
Module 1
Motion detection and respective human activity monitoring

    System Deployment
          Multi-Camera Registration & Deployment
          HotSpot Marking Utility
          System GUI

    System Initialization
          Motion Detection
          MER Formation

    Multi-Threading
Module 1 (cntd..)
Motion detection and respective human activity monitoring

Critical Difficulties Encountered– Camera Zoom [View Angle , Shadows]




        Camera Mounted in Room            Camera Mounted in the Model
Module 1 (cntd..)
Motion detection and respective human activity monitoring


 Prototype Environments :
    Room (Hostel); Model House


 Image Capture Devices :
      CMOS Webcam; 640 x 480 pixels; Frame Rate: 30 fps; View Angle: 54 degrees


 Platform :
    .NET 2.0 based Aforge C# Framework (open source)
          Using the AForge.Imaging & AForge.Vision libraries
                                                  [http://code.google.com/p/aforge/]
Module 1 (cntd..)
Motion detection and respective human activity monitoring


 User Detection Approach :
    Current–Previous, Current–First Algorithm
    Blob Motion Algorithm


 Capturing Results :
    MER (Minimum Enclosed Rectangle) Drawing    (.NET Imaging Library)
    Weighted Average, K-Means/K-Means ++ Algorithm


 Activity Monitoring :
                                                  (.NET Imaging Library)
    Initial Hotspot outline by the User on the S/W
    Hotspot & MER Overlap Detection using Crossing Number Method
Module 1 (cntd..)
Motion detection and respective human activity monitoring
                                           Hotspot marking by user




    MER formation for Motion Detection
  Module 1 (cntd..)
  Motion detection and respective human activity monitoring




                                                     Multi-Cam Supportive GUI



User-friendly Camera Registration        Spawning Threads for Concurrent Multi-User Support
System Architecture




                      MODULE 2
  MODULE 2
Module 2
Decision taking software engine – an expert system

    Knowledge Base Design


    Action Triggering
           Overlap Calculations
           Other Parameter Calculations

    Reaction Generation


    Serial Transmission
Module 2
Decision taking software engine – an expert system


 The brain of the system– responsible for the actual “monitoring”
      Decides when to turn off what


 Expert System :
    Defining classes for rules
    Having defined facts in a relational database
    Using inference techniques to judge which rules, based on user activity, stand, and
     hence which devices are to be respectively switched off
    Rules and Facts subject to changes and modifications based on user behavior


 Overview :
          Small scale expert system with a dynamically adjusting knowledge base
System Architecture




                      MODULE 3
  MODULE 3
Module 3
Serially interfaced appliance control mechanism


 Connection to Devices :
    Serial port communication via RS232
    Microcontroller
    Interface circuitry
    Device Model House.




               RS232


    Computer           Microcontroller     Interface   Electronic Devices
Module 3
Serially interfaced appliance control mechanism




           Trainer Board Circuitry                Vero Board Circuitry
Module 3
Serially interfaced appliance control mechanism


        Signal serially transmitted by the
        software



             Signal interpreted by the
             microcontroller




                 Devices switched on/off
Module 3
Serially interfaced appliance control mechanism

 Serial Communication
    Data is transmitted to the microcontroller via RS-232 standard
    DB- 9 connector is used to connect to the microcontroller
 Hardware circuitry
    ATMEL 89C51
    MAX 233
    2n222 Transistors
    Diodes
    6 volt relays
    Capacitors
    Resistors
Module 3
Serially interfaced appliance control mechanism


Circuit : Part I
Module 3
Serially interfaced appliance control mechanism


Circuit : Part II (implemented on Veroboard)
Module 3 (cntd..)
Serially interfaced appliance control mechanism


 Microcontroller ( ATMEL 89C51):
      Receives data serially from computer port
      Processes the data
      Sends data to its port to which devices are connected
      Turning relays on and off

 MAX 233
      RS-232 not compatible with today’s microcontrollers
      Line driver to convert RS-232 signal to TTL logic levels


 Interface Circuit :
      BJT based Relay-Driver (2n222):
          It provides sufficient current to drive the relay
      Diode :
          When the relay is switched off, magnetic field generated inside produces a high voltage, which can damage
           circuit
          Diode prevents short circuiting when the relay is switched off
Module 3 (cntd..)
Serially interfaced appliance control mechanism


 Interface Circuit (cntd..) :
    Resistor :
         It prevents flow of high current from the transistor to the ground on the
          application of VCC
    Relay :
         Mechanical relay is used to switch device off
Modeling the Prototype
The Model House (realistic cut-scale)


                                                                                    Outer
                                  ROOM
                                                                            Slide-Up-and-Removable
                                          Normal Walls
                                                                                    Walls



                  Wiring concealed in between by the removable Outer Wall


 Model Specifications:
    3.5‘ x 3.5‘ x (equivalent height)
    Two bedroom house (1.5‘ x 1.5')
    Double-walled structure with the outer walls removable
    The house is raised     1 feet in height, by supports, for magnet movements
    Test Men with magnets are used to depict movements in the household
Modeling the Prototype (cntd..)
The Model House




                  House Dimensions: 3.5’ x 3.5’ x 2’   Room Dimensions: 1.5’ x 1.5’ x 8”
Modeling the Prototype (cntd..)
The Model House with Domestic Appliances Installed




      Installed Appliances: Room Lights, Room Fan, Bed-side Lamp & Study-table Lamp
(Work-based) Modular Completion Status
 Module 1 : Motion detection and respective human activity monitoring
      System Deployment
             Multi-Camera Registration & Deployment
             HotSpot Marking Utility
             System GUI
      System Initialization
             Motion Detection
             MER Formation
      Multi-Threading (debugging issues!)

 Module 2 : Decision taking software engine – an expert system
      Action Triggering
              Overlap Calculations
              Other Parameter Calculations
      Knowledge Base Design
      Reaction Generation
      Serial Transmission

 Module 3 : Serially interfaced appliance control mechanism
      Trainer/Vero-board Circuit Designing
      Microcontroller Interpretation
      Electrical Interfacing of Appliance
      System Optimization
Envisioning Ahead

 Microsoft Imagine Cup 2008




 GIKI Participation @ Different Technology Competitions ~ Softec ’08
 Submissions on Various Research-based Platforms for Domotics
 Video Documentation

				
DOCUMENT INFO
Shared By:
Categories:
Tags:
Stats:
views:0
posted:3/26/2013
language:Latin
pages:35