Measurements of Japanese cedar growth on rough surface using spectral imaging analysis

Abstract

Abstract Accurate measurement of tree-ring increments on rough wood surfaces in the field is essential for evaluating forest productivity and merchantability. Conventional imaging approaches are hindered by intensity non-uniformity across the cross-section, particularly between the darker heartwood in the central region and the lighter sapwood towards the periphery. This study presents a spectral imaging method designed to compensate for intensity non-uniformity across the wood cross-section. By equalizing intensity variations between heartwood and sapwood regions, the method enables reliable identification of annual growth rings on unprocessed, rough surfaces of Japanese cedar. The spectral imaging station and the procedure for acquiring spectral images are described. Spectral characteristics that enhance ring visibility are analysed, and an algorithm is proposed for selecting the optimal spectral range to maximise contrast in the intensity-balanced image. The method was applied to 20 rough-surfaced Japanese cedar samples imaged at a spatial resolution of 0.05 mm/pixel. Annual ring increments were successfully quantified on these samples. Although validation on a larger and more diverse set of samples is required to confirm robustness and versatility, the key contribution lies in the proposed spectral imaging technique, which effectively addresses intensity non-uniformity and enables reliable dendrometric measurements on challenging field samples.