Show simple item record

The Weakly Identifying System for Doorway Monitoring

dc.contributor.advisor Ellis, Carla
dc.contributor.advisor Brady, David
dc.contributor.advisor Cox, Landon
dc.contributor.advisor Parr, Ron Jenkins, Christopher James 2007-05-10T15:22:30Z 2007-05-10T15:22:30Z 2007-05-10T15:22:30Z
dc.description.abstract The System Architecture for Tracking Individuals (SAFTI) is an indoor person location tracking system designed for use in the field of pervasive computing. SAFTI provides location tracking in environments where cameras are too privacy invasive, where tracking devices are too costly, insecure or inconvenient, and where usability is a high priority. While many location tracking systems satisfy each of these constraints individually, SAFTI satisfies all three constraints simultaneously. Upon entering and exiting SAFTI buildings, users submit identification credentials. Once inside the building, using SAFTI is effortless - simply passing through doorways is sufficient for supplying SAFTI with the information it needs to perform location tracking. An integral part of SAFTI is the Weakly Identifying System for Doorway Monitoring (WISDOM). These instrumented doorways contain a variety of infrared, ultrasonic and pressure sensors that detect the direction of passage and measure each user's body size and shape. We quantify the measurement and identification accuracy of WISDOM by analyzing data collected from a user study containing 530 passes through a WISDOM prototype from 10 different subjects. We combine the results from WISDOM with large publicly available anthropometric databases to evaluate how accurately SAFTI performs location tracking with respect to building size, density of occupants, and matching algorithm used.
dc.format.extent 1148162 bytes
dc.format.mimetype application/pdf
dc.language.iso en_US
dc.subject pervasive computing
dc.subject biometrics
dc.subject location
dc.subject tracking
dc.subject ambient
dc.subject recognition
dc.title The Weakly Identifying System for Doorway Monitoring
dc.type Dissertation
dc.department Computer Science

Files in this item


This item appears in the following Collection(s)

Show simple item record