Title :
Aura Removal Algorithm for High-Temperature Image Quantitative Analysis Systems
Author :
A. Fabijanska;D. Sankowski
Author_Institution :
PhD student, head of the Computer Engineering Department, Technical University of Lodz, Stefanowskiego 18/22, 90-924 Lodz, Poland. E-mail: an_fab@kis.p.lodz.pl
fDate :
6/1/2007 12:00:00 AM
Abstract :
This paper describes high temperature image quantitative analysis systems. The specificity of measurements based on images presenting heat-emitting objects is considered. Particular attention is paid to computerised system for surface properties determination. A brief description of the system is given. The measurement process is presented. Problems connected with processing images of hot objects emitting thermal radiation are deliberated. Especially aura phenomena and their influence on image segmentation quality are considered. Authors´ method of aura removal is proposed. Results of applying the algorithm to exemplary high temperature images are presented. Removing aura from analysed image strongly improves quality of object shape extraction due to image segmentation process.
Keywords :
"Image analysis","Algorithm design and analysis","Image segmentation","Layout","Digital images","Temperature measurement","Computer vision","Equations","Shape","Data mining"
Conference_Titel :
Mixed Design of Integrated Circuits and Systems, 2007. MIXDES ´07. 14th International Conference on
Print_ISBN :
83-922632-4-3
DOI :
10.1109/MIXDES.2007.4286236