Strategy for grading natural materials using a two-step classification procedure

Bjarne Kjaer Ersboll; Knut Conradsen

doi:10.1117/12.145549

28 May 1993 Strategy for grading natural materials using a two-step classification procedure

Bjarne Kjaer Ersboll, Knut Conradsen

Proceedings Volume 1821, Industrial Applications of Optical Inspection, Metrology, and Sensing; (1993) https://doi.org/10.1117/12.145549
Event: Applications in Optical Science and Engineering, 1992, Boston, MA, United States

Abstract

Quality grading of natural materials such as fur, wood and cork has traditionally been done manually. Since the grading by humans is done by assessing the visual quality of the material it has been tempting to try to automate the process using some kind of image analysis. Much research has been devoted to solving this problem. Grading rules for natural materials usually consist of a description of which types of defects are allowed for each grade and how severe they may be. The complexity of these grading rules and the variability of natural materials taken into account makes expert system or artificial intelligence approaches infeasible. This paper proposes a method for automatic grating of natural materials using statistically based classification techniques. The method involves two classification steps: in the first step a pixel wise classification of defects is performed based upon a number of features generated from the original image; the second step utilizes the relative distribution of defects to perform a final grading assessment. The method is illustrated using the grading of beech wood slabs as an example.

Citation Download Citation

Bjarne Kjaer Ersboll and Knut Conradsen "Strategy for grading natural materials using a two-step classification procedure", Proc. SPIE 1821, Industrial Applications of Optical Inspection, Metrology, and Sensing, (28 May 1993); https://doi.org/10.1117/12.145549

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available