You’ve Got SARS!!

1. You’ve Got SARS!! Group 6 Brent Anderson Lauren Cutsinger Martin Gilpatric Michael Oberg Matthew Taylor Capstone Spring 2006

2. Presentation Outline • Project Overview • Implementation • Hardware • Software • Project Organization • Risk and Contingency

3. IR in Medical Applications • First-Pass Diagnostic Tool • Soft tissue injuries • Circulatory Problems • Research • Biological response to low frequency appliances • Physical exercise and hypothermia research • Passively screen for a selection of biological agents • Quick response to possible pandemics • Airports, office buildings and other mass transit areas • SARS detection in several airports • SARS begins with a fever greater than 38°C (100.4°F)

4. Project Overview • Design an infrared tracking system that will control a motorized camera platform. • Track infrared image of person. • Display IR image. • Determine temperature of person for possible disease detection.

5. Thermopile Temperature Sensors • Multiple thermocouples in series that provide an extremely sensitive IR sensor. • Bulky and expensive compared to other types of IR sensors.

6. So Why Thermopiles? • Breakthroughs in Manufacturing process. • Lower cost • Much smaller footprint • Very accurate at –50 to 450 ° C • Correct spectrum for Heat • Comes in a 32x32 array!

7. System Overview SPI PWM SPI Serial SPI Major Components • IR Camera • PIC Processors • Camera Mount • Motors • PCB • Output (PC) VGA

8. Hardware

9. Why PIC18F4550 • Harvard Architecture • Simultaneous data and program memory access • Built in SPI interface • Camera • PIC communication • Two PWM Channels for Stepper Motors • Development Environment • On Chip Programmer • 48MHz clock • 2 UARTs for computer communication • USB 2.0 Interface (optional)

10. Why ATX25C? • Thermopile 32x32 Array • Low Resolution allows for faster data crunching. • Provides enough data for human detection. • SPI Interface • Compatible with PIC18F4550

11. Why Stepper Motors • High precision • Up to 1.5 degrees of accuracy • Compatible With PIC18F • PWM signals • Low cost for precision • Easy positioning

12. Software Outline • Tracking Algorithms • “Hot Spot” Detection • Edge Detection • Motor control • On Chip Pulse Width Modulation (PWM) • Display • Comunication with a PC for display and control.

13. Targeting Software Goals • Identify the portion of the picture containing the face of the subject.

14. Tracking DemonstrationAveraging • Averages are taken along both the horizontal and vertical axis. • The highest value of these averages indicates the hottest cross section in each dimension. • The intersection of these cross sections should be the center of the warmest body. In this case we are hoping that this body will be the face of the subject.

15. Tracking DemonstrationDifferentiation • By differentiating the cross section averages generated in the previous method these plots can be generated. • The peaks of these plots indicate the planes where there has been the greatest change in value. • This method will find the boundries of what should be the head of the subject.

16. Interface • Motor control • Interacting with onboard PWM for use with motor control logic. • Display data via a PC • UART connectivity with a Linux machine to display pertinant information and raw data. • Possibility of screen capture. • Possibility of future additions though PC based processing.

17. Project Goals and Constraints • Baseline Objectives • Advanced Goals • USB2 • Advanced Display and User Interface • Printing • Constraints and Contingency Plans

18. Baseline Objectives • IR Camera <-> SPI bus w/ two µcontrollers • Real-time processing of incoming data stream at 10 Frames Per Second • Dynamic image tracking • Determine and record temperature data • UART Data stream to PC

19. Baseline Objectives (Cont) • PC Client software for User Interface and Control • Display of RAW Camera Data • False Color Representation with Heat Distribution • Summary Statistics and Indication of Abnormal Temp

20. Advanced Goals and Extensions • USB2 for data transmission, using UART for control • Improvements to PC Software • Image Interpolation • Additional Controls • Pause • Screenshot • Print Screen

21. Project ConstraintsAirport Usage • Tracking a single hot-spot (no crowds) • Limits on tracking speed. • Target must be within predefined distances (lens specific) • PERSON MUST WEAR CLOTHING!!!! • Face must be exposed. • Heavy clothing preferred.

22. Hardware Constraints • Timing data on the SPI bus for communication between two processors • UART speeds. • Camera Resolution • Will 32x32 resolution provide accuracy? • Image Processing Speed • Can we processes 10 FPS • Smooth Motor Control

23. Risks and Contingencies • Parts Delivery • IR Camera: Can use CCD which supports SPI • Difficulties with optics • Software workarounds / mods to core algorithms • Replacement optics • Inability to process data flow in real-time • Slow the data, lower the frame rate • Stage the data with external RAM components • Add additional micro-controllers and design parallel processing

24. Team Planning • Cost of Implementation • Individual Member Tasks • Team Milestones and Goals • Questions?

25. Costs

26. Tasks

27. Timeline

28. Questions?