DocumentCode :
2893310
Title :
Plain Ceramic Tiles Surface Defect Detection Using Image Processing
Author :
Shire, Atul N. ; Khanapurkar, M.M. ; Mundewadikar, Rajashri S.
Author_Institution :
Electron. Eng. Dept., G.H. Raisoni Coll. of Eng., Nagpur, India
fYear :
2011
fDate :
18-20 Nov. 2011
Firstpage :
215
Lastpage :
220
Abstract :
In the ceramic tile industry bulk amount of ceramic tiles are manufactured, it is very difficult to monitor the quality of each and every tiles manually. Lot of human resources are required for the defect detection of the tiles. Also it is quite tedious and time consuming method. Considering this criteria, an automated defect detection and classification technique has been proposed in this report that can have ensured the better quality of tiles in manufacturing process as well as production rate. Our proposed method plays an important role in ceramic tiles industries to detect the defects and to control the quality of ceramic tiles. This automated classification method helps us to monitor the defects within a very short period of time and also to decide about the recovery process so that the defected tiles may not be mixed with the good quality tiles.
Keywords :
ceramic industry; image classification; image enhancement; image processing; quality control; tiles; automated defect classification technique; automated defect detection technique; ceramic tile industry; human resources; image processing; plain ceramic tile surface defect detection; quality control; recovery process; tile manufacturing process; Ceramics; Image color analysis; Image edge detection; Low pass filters; Maximum likelihood detection; Nonlinear filters; Tiles; Defected Ceramic Tiles; Fresh Tiles; Image enhancement; Preprocessing;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Emerging Trends in Engineering and Technology (ICETET), 2011 4th International Conference on
Conference_Location :
Port Louis
ISSN :
2157-0477
Print_ISBN :
978-1-4577-1847-2
Type :
conf
DOI :
10.1109/ICETET.2011.63
Filename :
6120585
Link To Document :
بازگشت