Title
A Method of Industrial Parts Surface Inspection Based on an Optics Model
Abstract
This paper proposes a new approach based on an optics model for highly reliable surface inspection of industrial parts. This method uses multiple images taken under different camera conditions. Phong's model is employed for surface reflection and then the albedo and the reflection model parameters are estimated by the least squares method. Experimental results show the proposed method's advantages over the conventional binarization method for easily determining a threshold of product acceptability and coping with changes in light intensity when detecting defects.
Year
DOI
Venue
1998
10.1007/3-540-63930-6_122
ACCV (1)
Keywords
Field
DocType
optics model,industrial parts surface inspection,least square method
Least squares,Computer vision,Surface reflection,Computer science,Optics,Image processing,Albedo,Artificial intelligence
Conference
ISBN
Citations 
PageRank 
3-540-63930-6
2
0.61
References 
Authors
5
4
Name
Order
Citations
PageRank
Norifumi Katafuchi151.41
Mutsuo Sano24211.24
Shuichi Ohara3212.51
Masashi Okudaira4182.20