The automatic classification of surface defects on wood boards is need
ed in respect to productivity and accurate grading. Thus the recogniti
on of knots, one of the defects in wood, has been studied. An image pr
ocessing technique was used for the recognition of knots in boards. Th
is technique was found to be useful because the color of knots is defi
nitely different from that of the surrounding wood tissues. But knots
cannot be distinguished by color from the clear portions of lumber whe
n knot color is similar to the color of the heartwood. To solve this p
roblem, knots of hinoki (chamaecyparis obutusa Endl.) boards were exam
ined for differences of gloss between irradiation parallel and perpend
icular to their longitudinal directions. Also, the possibility of iden
tifying knots by the differences of gloss between those irradiated par
allel and perpendicular to the longitudinal direction of the board sur
faces of 11 species of conifers were examined. Boards planed by a sing
le surface planer and by a fixed knife planer were used in the study.
As a result, in all of the species having the same features of gloss,
the knots were recognized from the surrounding trunk wood. However, it
was difficult to identify the knots in cases when the gloss of boards
was affected by surface quality. This indicates that the machinabilit
ies of wood species, as well as the procedure of processing boards, ha
ve an influence on the possibility of recognizing knots by differences
of gloss.