Development of Video Image Detection Algorithm for Smoke Plumes
ProQuest, 2007 - 77 pages
The objective of this work is to develop a predictive activation time algorithm for smoke plumes for axonX's SigniFire video image detection (VID) system from recent tests in a large volume space performed at the University of Maryland's (UMD) Cole Field House. The SigniFire system was able to detect a smoke plume from distances of 37.8 m (124 ft) to 54.9 m (180 ft) typically before the smoke reached the ceiling. The goal is to establish an understanding of the significant parameters affecting activation time based on observations and trends from the video image data. As a result of the understanding, insight into a predictive algorithm is developed, which is the first step toward future use of the VID system for a performance based design.
What people are saying - Write a review
We haven't found any reviews in the usual places.
Development of Predictive Algorithm
ANOVA arena floor AxonX calculated Camera Alarm tactivation CCTV coefficient Cole Field House Contrast Ratio correlation Density at 1.19 Density at 2.72 detection system Edwards System Technology Eight Cameras emitter at High emitter at Low fire detection fire sources flaming fires foot-candles fuel parameter heat release rate height heptane Heskestad high lighting level instrumentation apparatus lighting conditions low lighting level luminance machine vision mass entrainment rate nuisance sources number of pixels optical density optical density measurements Outline Growth Rate overall P-value performed Photometric Measurements taken photometric survey plume algorithm plume and background plume mass entrainment Plume Outline Growth prediction presented as Equation radiant heaters regression analysis SFA system SigniFire system Smoke 15 smoke algorithm smoke detectors smoke emitters smoke plume Statistical Output test series Thermocouple Thermocouple Temperatures total number Volume Sensor volumetric flow rate volumetric plume VSD-8 system