We present a method for field use to differentiate a natural mutation, curly birch (CB) (Betula pendula var carelica Sok.) from the ordinary silver birch (SB) (Betula pendula Roth.). Differentiating the two types is crucial for forest owners, since CB commands a high price, and the faster-growing SBs overshadow these trees. This creates a need for an automated, easily applicable method to differentiate between them. The presented device segregates intact trunks by recording the attenuation of 100 kHz ultrasonic longitudinal tone bursts launched radially through the trunks. We achieved 90% probability of correct characterization (0% probability of false positive) for 20 trunks.
We report on computerized differentiation of two birch types. Curly birch (CB, Betula pendula var. carelica Sok) commands a higher price than a normal silver birch (SB, Betula pendula Roth). Hence it is crucial to differentiate the two wood types when the trees are young. We studied the possibility to use ultrasound for such differentiation. A propagation velocity of 4MHz longitudinal tone bursts, transmitted through block samples of 20×20mm2 cross section, comprising of both CB and SB woods, was determined. The samples originating from southern Finland were sawed so that the sound propagation direction was longitudinal or radial with respect to the trunk. One sample set comprised of seven different sample thicknesses with a range of 2–12mm. From the time-of-flight measurements of the samples (19±1%weight humidity) the wave propagation velocity under laboratory conditions (50±5%RH, 23±1°C) was determined from a least-squares fit. The results indicate a significant difference (t-test p=0.032 and velocity difference of 24±8%) in longitudinal direction and a highly significant difference (t-test p=0.001 and velocity difference of 22±10%) in the radial direction between the two wood types. A blinded probability of detection test was conducted using 48 samples originating from two different trunks as well as from single trunks’ curly and noncurly sections. The results indicate 93% probability of correct type classification using computerized clusterization.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.