Title of article :
A survey of hierarchical non-linear medical image registration
Author/Authors :
Arridge، Simon R. نويسنده , , Lester، Hava نويسنده ,
Issue Information :
روزنامه با شماره پیاپی سال 1999
Pages :
-128
From page :
129
To page :
0
Abstract :
A program to assess the impact of smoke on digital instrumentation and control (T & C) safety systems began in 1994, funded by the US Nuclear Regulatory Commission Office of Research. Digital I & C safety systems are likely replacements for todayʹs analog systems. The nuclear industry has little experience in qualifying digital electronics for critical systems, part of which is understanding system performance during plant fires. The results of tests evaluating the performance of digital circuits and chip technologies exposed to the various smoke and humidity conditions representative of cable fires are discussed. Tests results show that low lo moderate smoke densities can cause intermittent failures of digital systems. Smoke increases leakage currenls between biased contacts, leading to shorts. Chips with faster switching times, and thus higher output drive currents, are less sensitive to leakage currents and thus to smoke. Contact corrosion from acidic gases in smoke and stray inductance or capacitance are less important contributors to system upset. Transmission line coupling was increased because the smoke acted as a conductive layer between the lines. Permanent circuit damage was not obvious in the 24 h ofcircuit monitoring. Test results also show that polyurethane, parylene, and acrylic conformal coatings arc more effective in protecting against smoke than epoxy or silicone. Common sense mitigation measures are discussed. Unfortunately we are a long way from standard tests for smoke exposure that capture the variations in smoke exposure possible in an actual fire. © 1999 Elsevier Science S.A. All rights reserved.
Keywords :
Non-linear registration , Landmark-based splines , Elastic warping , Fluid deformation , Anisotropic scale space
Journal title :
PATTERN RECOGNITION
Serial Year :
1999
Journal title :
PATTERN RECOGNITION
Record number :
14365
Link To Document :
بازگشت