Design and Performance of an Image Processing Occupancy Sensor

Anthony Florita

Research output: Contribution to conferencePaper


The design and performance of a novel image-processing occupancy sensor (IPOS) based on production-grade embedded hardware are described. The IPOS overcomes several issues related to traditional passive infrared (PIR) and ultrasonic-based motion sensors. It also provides a rich set of feedback signals that more sophisticated applications can use for occupant-optimized lighting, daylighting,temperature,and ventilation control, as well as with building security systems. The IPOS was tested as part of an integrated system consisting of a building energy management system (EMS) alongside a pair of PIR sensors produced by different manufacturers. IPOS output was fed to the EMS via the building automation and controls network (BACnet) and logged alongside the PIR outputs in an officeenvironment over several days. Tests suggest the IPOS device identifies occupancy events better than two commercial PIR sensors (94.1% accuracy compared to 73.1% on average). Also, IPOS could correctly identify stationary occupants whereas the PIR sensors could not.
Original languageAmerican English
Number of pages8
StatePublished - 2012
EventSecond International Conference on Building Energy and Environment 2012 - Boulder, Colorado
Duration: 1 Aug 20124 Aug 2012


ConferenceSecond International Conference on Building Energy and Environment 2012
CityBoulder, Colorado

NREL Publication Number

  • NREL/CP-5500-54800


  • built environment
  • image processing
  • lighting
  • occupancy sensing
  • ventilation


Dive into the research topics of 'Design and Performance of an Image Processing Occupancy Sensor'. Together they form a unique fingerprint.

Cite this