Color grading of beech parquet blocks by using spectral data

摘要

Modem-day wood industry has started to apply machine vision and other automation technology in grading raw material and intermediate and final products. Color grading is more or less based on the human visual sense, but color grades can be standardized and the grading process could be made more accurate and faster by automating the process. In the present study, the objective was to develop and test the CLAFIC (CLAss Featuring Information Compression) and Mahalanobis (based on the Mahalanobis Distance) classifiers using spectral information that automatically grades the beech parquet blocks into five pre-specified commercial grades. The results were compared with the results of manual grading made by an expert from a parquet factory. The classifier results correlated very well with the manual results. Some possibilities for improving the automatic color grading systems are discussed.