Wood Identification Research Articles

Predicting hardwood porosity domains: Toward cascading computer-vision wood identification models

Prior work on computer-vision wood identification (CVWID) for North American hardwoods yielded two independent deep learning models – a 22-class model for diffuse-porous woods and a 17-class model for ring-porous woods – but did not address semi-ring-porous woods nor provide a CVWID solution for an unknown specimen without a human first determining which model to deploy. As untrained human operators would lack the anatomical proficiency to differentiate among porosity domains, it is necessary to develop a consolidated model that can identify diffuse-, ring-, and semi-ring-porous woods. Previous research suggests that prediction accuracy might decrease as class number grows. A potential strategy to reduce the number of classes a CVWID system must consider at a time is to hierarchically deploy a cascade of models. In pursuit of a unified model that can cover North American hardwoods of all porosity types, this study compared the accuracies of a consolidated 39-class (ring- + diffuse-porous) model and a consolidated 42-class (ring- + diffuse- + semi-ring-porous) model with a two-tiered, cascading model scheme whereby images are first differentiated into three porosity domain classes and then again into only those taxonomic classes with that porosity. The results showed that the cascading model scheme can mitigate the accuracy reductions incurred by the 42-class model and nearly eliminate the occurrence of cross-domain misidentifications.

Sustainable Lignin Extraction Using a Mild Acidic γ-Valerolactone Process for the Identification of White and Red Oaks.

Current wood identification struggles to differentiate white and red oak (Quercus alba and Quercus rubra) due to highly similar microstructures, as demonstrated by morphological microscope analysis. The present research explores lignin composition as a potential discriminating factor. Here, a rapid and sustainable method for extracting high-quality lignin from oak samples using acidic γ-valerolactone (GVL) under mild conditions is described. As-extracted lignin is thoroughly characterized using various analytical methods, and results reveal a distinct structural difference between the lignin from the two species. White oak lignin possesses a unique "Hibbert ketone" unit detectable by nuclear magnetic resonance spectroscopy (NMR), which is absent in red oak lignin. In addition, infrared spectroscopy differentiates the species based on specific carbonyl groups present in their lignin. These findings suggest that identifying the presence of the Hibbert ketone unit in lignin may offer a highly efficient and reliable method for differentiating white and red oak, opening new avenues for wood identification.

Wood discrimination of six commonly traded Phoebe and Machilus species using high-resolution plastid and nuclear DNA barcodes

Abstract Phoebe and Machilus trees, whose wood is known as Nanmu in China, face endangerment owing to long-term manmade destruction and climate change. Therefore, effective and accurate forensic tools to identify Nanmu species are required. This study developed a method for DNA identification of woods from six commonly traded Phoebe and Machilus species. DNA was efficiently extracted from oil cell-rich wood tissues, DNA barcodes were screened, species were compared, and a Nanmu DNA database was constructed. Comparison and analysis involved 26 sets of plastid genomes of Phoebe and Machilus. The five highest variable regions (Pi>0.009; psbA-trnH, psbJ, psbJ-petA, rpl32-trnL, and ycf 1) were identified, and three nuclear regions (ITS, LEAFY, and RPB2) were compared. Among single DNA fragments, the nuclear DNA region LEAFY proved optimal for identifying Nanmu species, while psbA-trnH+psbJ+ycf1 was the optimal plastid combination locus. Compared with distance-based (TaxonDNA) and machine-learning analyses, the tree-based method had a lower success rate for species identification. Four labelled Nanmu samples were selected to validate the method and clustered with P. zhennan and P. sheareri data based on the LEAFY region and plastid combination locus. These findings will contribute to the conservation, classification, architecture maintenance, and trade monitoring of Nanmu natural resources.

Interlaboratory and cross-platform accessibility of time-of-flight wood identification database

The mass spectral database of tree species built by US Fish and Wildlife Service has thousands of entries and has been a valuable resource to combat illegal logging and international trade. The database was and continues to be constructed using a particular ambient ionization time-of-flight mass spectrometry (TOF-MS) platform in the agency branch in Ashland, OR, with which queries of unknown wood samples are investigated exclusively. Laboratories that operate different MS instruments also have an interest in using the database if they can produce valid matches to known samples compatible with the database. Four species were selected for inter-laboratory comparison using Orbitrap MS instruments and the equivalent TOF-MS platform with direct analysis in real time ionization of institution-sourced wood samples. Identities of the known samples were confirmed by examination of their microscopic wood anatomy. Orbitrap analysis was able to identify each species as confidently as the TOF instruments, often with less variation in spectra but not necessarily greater mass accuracy or better-matched signal abundance to the control database. The Orbitrap program also had to be doubled to two scanned mass ranges appended for greater peak intensity, before spectra could be correctly matched to the database, but the program was successful.

Uniseriate ray characteristics for wood identification and quality indices of six Korean oak species

Uniseriate ray characteristics for wood identification and quality indices of six Korean oak species

Robustness of a macroscopic computer-vision wood identification model to digital perturbations of test images

Summary Distribution shift, a phenomenon in machine learning characterized by a change in input data distribution between training and testing, can reduce the predictive accuracy of deep learning models. As operator and hardware conditions at the time of training are not always consistent with those after deployment, computer vision wood identification (CVWID) models are potentially susceptible to the negative impacts of distribution shift in the field. To maximize the robustness of CVWID models, it is critical to evaluate the influence of distribution shifts on model performance. In this study, a previously published 24-class CVWID model for Peruvian timbers was evaluated on images of test specimens digitally perturbed to simulate four kinds of image variations an operator might encounter in the field including (1) red and blue color shifts to simulate sensor drift or the effects of disparate sensors; (2) resizing to simulate different magnifications that could result from using different or improperly calibrated hardware; (3) digital scratches to simulate artifacts of specimen preparation; and (4) a range of blurring effects to simulate out-of-focus images. The model was most robust to digital scratches, moderately robust to red shift and smaller areas of medium-to-severe blur, and was least robust to resizing, blue shift, and large areas of medium-to-severe blur. These findings emphasize the importance of formulating and consistently applying best practices to reduce the occurrence of distribution shift in practice and standardizing imaging hardware and protocols to ensure dataset compatibility across CVWID platforms.

Evaluation of non-anatomical characteristics for wood identification of six Korean oak species

Summary This study aimed to investigate and compare the non-anatomical characteristics of six Korean oak wood species (Quercus variabilis, Q. serrata, Q. mongolica, Q. dentata, Q. aliena and Q. acutissima) to provide identification keys and quality indices for the effective utilization of these species. Non-anatomical characteristics, including heartwood fluorescence, fluorescence and color of water and ethanol extracts, froth test, chrome azurol S test and burning splinter test, were evaluated according to the International Association of Wood Anatomists (IAWA) lists. None of the six Korean oak species displayed heartwood fluorescence or fluorescence in the ethanol extract. Only Q. serrata exhibited bright blue fluorescence in the aqueous extract. Each species showed distinct colors in both water and ethanol extracts, and the water extract of all species was darker than that of the ethanol extract. A weak positive reaction in the froth test was observed only for Q. serrata. Furthermore, Q. variabilis and Q. acutissima showed positive reactions to both heartwood and sapwood in the chrome azurol S test. The burning splinter test revealed that only Q. acutissima was transformed into charcoal. Thus, water extract fluorescence and color, ethanol extract color, froth test, chrome azurol S test, and burning splinter test can be used to identify the six Korean oak species.

New fossil woods from the middle Eocene climate optimum of north-central Turkey

The middle Eocene climate optimum, a crucial greenhouse event in Earth’s life history, occurred approximately 40 Ma. Fossil wood identifications of middle Eocene age reveal important evidence for the terrestrial tropical/subtropical conditions during this period. The purpose of the present study is to identify new fossil woods from the middle Eocene of north-central Turkey and to evaluate the paleoclimate and paleovegetational history of the north-central Turkey based on new records. Seven fossil woods were collected from the middle Eocene Göynücek Volcanics of the Çekerek Formation (Sarıkaya village, close to Çekerek, Yozgat). Two new species Apocynoxylon umut-tuncii Akkemik and Mantzouka, n. sp., and Ficoxylon anatolica Akkemik and Mantzouka, n. sp., are described, together with two more coniferous species Pinuxylon cf. P. tarnocziense (Tuzśon, 1901) Greguss, 1954 and Juniperoxylon cf. J. acarcaeaAkkemik, 2021a. Although dealing with a rather low number of fossil wood findings, indistinct boundaries of the tree rings, low xeromorphy ratios and high conductivity values in these two new fossil angiosperm species and less visible and indistinct boundaries in the two fossil coniferous species may reflect the warm, humid tropical conditions of the middle Eocene of north-central Turkey. The results are consistent with our former paleoecological evaluation based on the fossil woods including Actinodaphnoxylon zileensis Akkemik and Mantzouka in Akkemik et al., 2021 from the same geological unit and Palmoxylon sabaloidesGreguss, 1969 from the same area.

DNA Barcodes for Wood Identification of Anatomically Similar Species of Genus Chamaecyparis

DNA Barcodes for Wood Identification of Anatomically Similar Species of Genus Chamaecyparis

The application of chromatography and complementary techniques for wood identification — State of the art and future directions

Summary Wood identification, whether it involves distinguishing between known species or identifying unfamiliar wood samples, is a scientific field increasingly valuable across various disciplines ranging from biology to criminology, structural engineering, and art conservation. It carries a growing economic, commercial, social, and ecological significance. Numerous scientific methods have been employed for wood identification. Visual analysis techniques, such as macroscopy, optical microscopy, scanning electron microscopy, X-ray tomography, and computer-assisted wood identification, play a pivotal role. Analytical approaches like mass spectrometry, near-infrared spectroscopy, thermogravimetry, and DNA barcoding have also been utilised. However, chromatography stands out due to its exceptional precision in providing quantitative and qualitative results, prompting its development and widespread use. This paper offers a critical review of the role of chromatography in wood identification. Advantages are highlighted, including the capability to identify specific components unique to each wood species, thus achieving remarkable differentiation at the species level. Additionally, potential drawbacks are discussed, such as the time-intensive sample preparation procedures, mainly when dealing with materials like wood-polymer composites (WPC), and the absence of an open-source database. This comprehensive analysis aims to provide a broader perspective on the possibilities and limitations of this technique.

Chemical reagent for detecting tension wood in selected tree species

Reaction wood is a wood defect arising during the growth of the tree in the part of the trunk that is under tension (hardwood tree species) or compression (coniferous tree species). Beech (Fagus sylvatica L.) tension wood has different anatomical and chemical characteristics than normal (opposite) wood. The difference in density is conditioned by the percentage of the gelatinous layer (G-layer). Fibre cells in reaction beech wood have a different cell wall structure and a different chemical composition. Tension wood cannot be detected by the naked eye. It is only possible to assume its occurrence based on the macroscopic characteristics of the logs, such as a woolly surface, taper or eccentric pith, and so forth. However, these are imprecise and unreliable methods that have minimal effectiveness, especially when shortening the length of the log for cut-outs. This study aimed to create a unique chemical reagent for the detection of tension wood in logs and timber and wood products immediately. The present research can contribute to the mitigation of flaws resulting from the reaction of wood in timber production while addressing noticeable constraints in manufacturing, such as energy resources and the availability of wood raw materials. This can be achieved through the efficient identification of reaction wood in materials. The colour change is only temporary and will fade over time. After the chemical reagent has dried on the surface, the surface can be milled. The colour change extends to a depth of approx. of 3 to 5 mm.

A comparative study on the performance of terahertz, near-infrared, and hyperspectral spectroscopy for wood identification

ABSTRACT Wood species identification is of paramount significance in wood products manufacturing and applications. In contrast to traditional wood identification methods, spectroscopy-based technology offers a rapid, cost-effective, and efficient alternative. This study focuses on five wood species as experimental materials and aims to obtain three distinct wood spectra for each: near-infrared (NIR) spectra, hyperspectral image spectral information, and terahertz (THz) spectra. These spectra underwent pre-processing techniques such as Savitzky–Golay smoothing (SG), normalization, multiple scattering correction (MSC), and standard normalized variate (SNV), followed by dimensionality reduction through principal component analysis (PCA). Subsequently, the processed data were input into a partial least squares discriminant analysis (PLS-DA) for recognition. The results demonstrate the best recognition accuracy of 99.8% for THz spectra, 98.7% for NIR spectra, and 97.3% for hyperspectral image spectral information. The THz spectra exhibited the highest recognition accuracy, particularly with the SG-preprocessed THz spectra. These preprocessed spectra effectively removed noise and smoothed the spectral curves compared to the raw spectra.

Strength classes of brazilian hardwoods for structural design

For the elaboration of projects on timber structures, the Brazilian standard (ABNT, in Portuguese Associação Brasileira de Normas Técnicas) 7190 (ABNT 1997) ensures the correct application of physical-mechanical properties according to strength classes of lignocellulosic materials. This procedure eliminates the need for botanical identification of woods, since these strength classes support the efficient utilization of a wide range of woody varieties available in Brazil. Due to high mechanical resistances, the hardwoods are usually applied for structural projects of timber construction. This Brazilian standard document prescribes four strength classes (C20, C30, C40 and C60) for these woods, which are determined by characteristic value from the compressive strength parallel to the grain (fc0,k). But, these classes were obtained from experimental outcomes using a few wood varieties. The reorganization of strength classes should result in the best use of mechanical potentials of woods since the updating of these categories leads to the optimization of structural projects for timber construction in order to reposition this biomaterial at even more competitive levels. In this context, the present study aims to verify the current strength classes of hardwoods and if they lead to a good allocation of the fc0,k characteristic values. Otherwise, new strength classes can be determined for better allocations for efficient structural uses. As a result, 56 hardwoods were considered using 672 experimental determinations. Statistically, the current categories can lead to 12,5 % of incorrectly allocated values. The inclusion of C50 and C70 classes allow greater representation for these categories, in order to optimize the use of the hardwoods given the new strength classes. These findings ought to support future revisions of this Brazilian standard document.

Machine learning-based wood anatomy identification: towards anatomical feature recognition

Summary Computer vision-based wood identification has been successfully applied to recognize tree species using digital images of wood sections or surfaces. However, this image-to-species approach can only recognize a limited number of species due to two main reasons: 1) the lack of a good reference database requiring high-quality standardized images from multiple individuals of hundreds or even thousands of traded timber species, and 2) species not included in the reference database cannot be identified without expert knowledge. Another bottleneck is that the feature extraction process used by these species recognition approaches is a black box, thereby creating a discrepancy between machine learning features and wood anatomical features. This discrepancy prevents wood anatomists from understanding how these machine-learning algorithms work. Here, we survey currently existing methods used in feature extraction, classification, and deep learning methods applied in wood identification along with their pitfalls and opportunities. As an example of how the field could move forward, we launch the idea of building an image-to-features-to-species identification approach based on microscopic wood images as well as text files comprising wood anatomical descriptions. If we can manage machine learning-based algorithms to recognize the main wood anatomical traits that experts use to identify species in a (semi-)automated way, this would boost wood identification in two ways: (1) extensive reference databases for each species would become less crucial as the databases are ordered at the trait level, (2) timber identification would become more feasible for species that have not yet been included in the reference database as long as wood anatomical descriptions are available.

Discrimination of five commercial Guibourtia wood species using terahertz time domain spectroscopy combined with machine learning approaches

Abstract Terahertz waves hold significant potential for applications in wood identification, owing to their good penetration and distinctive fingerprints in wood. This study focuses on wood samples from five different Guibourtia species as the research objects. The terahertz time-domain spectroscopy (THz-TDS) is employed to acquire the spectroscopic signals of the wood samples and to extract their optical parameter data. The THz refractive indices are dimensionally reduced through principal component analysis (PCA), and three machine learning models, namely partial least squares-discriminant analysis (PLS-DA), random forest (RF), and support vector machine (SVM), are employed to classify the wood of five different Guibourtia species. Time delays of the wood samples from five different Guibourtia species are concentrated in the range of 60–62 ps and exhibit different amplitudes in the frequency domain. Refractive indices showed significant variations within the THz band. PCA for dimensionality reduction of terahertz time-domain spectral data significantly improves the recognition rate of machine learning models. Applying PCA to the refractive index data, the RF model achieves a highest recognition rate of 96.9 % and an overall classification accuracy of 98 %. Current results demonstrate that THz-TDS enables rapid, accurate, and non-destructive classification and identification of wood from the Guibourtia species.

CONSERVATION OF A PAINTED WOODEN COFFIN AT DAHSHUR ARCHAEOLOGICAL AREA

This paper aims to document the conservation processes of a polychrome wooden coffin in the Dahshur archaeological area dating back to the late period. The exterior part of the coffin is decorated with a painted layer. Visual observation, 2D Program, and Optical Microscopy (OM) were used. wood identification. The coffin was in a bad condition. It was covered with a thick layer of dust, losing parts of the painted and gesso layers, as well as other parts of these layers, were lost. Some parts were missing from the head area of the lid coffin. The conservation processes of the wooden coffin included mechanical and chemical cleaning, reattachment of the separated parts of the ground layer and painted layers, filling the edge of the painted layer, and consolidating the painted layer. The conservation process included mechanical cleaning using soft brushes, chemical cleaning using ethyl alcohol and distilled water for painting, stabilization of the separated gesso layer using Paraloid B72, filling cracks of the gesso layers using glass microballoon with Paraloid B72 and consolidating the painted layer with calcium oxide nanoparticles with Klucel G (hydroxypropyl cellulose) 0. 5%

The utilization of an olfactory machine in wood identification demonstrates a promising prospect: discerning disparities in emission profiles of volatile organic compounds between Picea abies and Pinus sylvestris

The utilization of an olfactory machine in wood identification demonstrates a promising prospect: discerning disparities in emission profiles of volatile organic compounds between Picea abies and Pinus sylvestris

Alternative identification of wood from natural fallen trees of the Lecythidaceae family in the Central Amazonian using FT-NIR spectroscopy

The scientific identification of natural fallen trees in tropical forests is complex due to the lack of fertile material in field collection. The study evaluated the use of near-infrared spectroscopy with Fourier-transform (FT-NIR) in the discrimination of wood from fallen trees of the Lecythidaceae family. Seven trees were collected in the Central Amazonian region (Brazil), from which 63 specimens were prepared from the wood, and NIR spectra were obtained on different wood surfaces (total 756 spectra). Chemometric models were developed with a spectral data set, and the Mahalanobis algorithm was applied. The discriminant model with 2nd derivative spectra improved the identification capacity, resulting in errors < 5% in the identification of genus Couratari (3 ssp.), Eschweilera (2 ssp.), Holopyxidium (1 sp.) and Lecythis (1 sp.). The comparison of the spectral signatures of samples of fallen trees and wood library revealed that even when wood was exposed to environmental weathering, around 50% of the original bands were preserved, favouring discrimination at the genus level. The accuracy of the chemometric models developed indicates the applicability of FT-NIR spectroscopy integrative in identifying fallen trees from the Lecythidaceae family in the tropical forests.

Qualitative and quantitative anatomical characteristics and radial variation of major cell components in Paulownia tomentosa wood grown in Korea

Qualitative and quantitative anatomical characteristics and radial variations of the major cell components in Paulownia tomentosa wood were examined using optical microscopy and X-ray diffraction to aid in wood identification and as quality indices. The vessel arrangement on the transverse surface was either ring-porous or semi-ring-porous. Most vessels had solitary pores, while some vessels had multiple radial pores. The axial parenchyma was generally confluent and partially of aliform type. Tyloses with high frequency in the vessel lumen and multiseriate rays (2 to 5 cells) were typical. The vessel diameter of earlywood and latewood was approximately 240 and 107 μm, respectively, with a range of 165 to 289 μm in earlywood and 55 to 149 μm in latewood. Ray height and fiber length were approximately 178 and 740 μm, respectively. The vessel diameter in both earlywood and latewood and the fiber length increased gradually with an increasing number of growth rings. Ray height was constant from the pith to the middle section and decreased toward the bark. The anatomical characteristics and radial variation of major components of P. tomentosa can be used as wood identification keys and quality indices.

Wood identification based on macroscopic images using deep and transfer learning approaches.

Identifying forest types is vital for evaluating the ecological, economic, and social benefits provided by forests, and for protecting, managing, and sustaining them. Although traditionally based on expert observation, recent developments have increased the use of technologies such as artificial intelligence (AI). The use of advanced methods such as deep learning will make forest species recognition faster and easier. In this study, the deep network models RestNet18, GoogLeNet, VGG19, Inceptionv3, MobileNetv2, DenseNet201, InceptionResNetv2, EfficientNet and ShuffleNet, which were pre-trained with ImageNet dataset, were adapted to a new dataset. In this adaptation, transfer learning method is used. These models have different architectures that allow a wide range of performance evaluation. The performance of the model was evaluated by accuracy, recall, precision, F1-score, specificity and Matthews correlation coefficient. ShuffleNet was proposed as a lightweight network model that achieves high performance with low computational power and resource requirements. This model was an efficient model with an accuracy close to other models with customisation. This study reveals that deep network models are an effective tool in the field of forest species recognition. This study makes an important contribution to the conservation and management of forests.