In recent years, due to the shortage of timber resources, it has become necessary to reduce the excessive consumption of forest resources. Non-destructive testing technology can quickly find wood defects and effectively improve wood utilization. Deep learning has achieved significant results as one of the most commonly used methods in the detection of wood knots. However, compared with convolutional neural networks in other fields, the depth of deep learning models for the detection of wood knots is still very shallow. This is because the number of samples marked in the wood detection is too small, which limits the accuracy of the final prediction of the results. In this paper, ResNet-34 is combined with transfer learning, and a new TL-ResNet34 deep learning model with 35 convolution depths is proposed to detect wood knot defects. Among them, ResNet-34 is used as a feature extractor for wood knot defects. At the same time, a new method TL-ResNet34 is proposed, which combines ResNet-34 with transfer learning. After that, the wood knot defect dataset was applied to TL-ResNet34 for testing. The results show that the detection accuracy of the dataset trained by TL-ResNet34 is significantly higher than that of other methods. This shows that the final prediction accuracy of the detection of wood knot defects can be improved by TL-ResNet34.

Taxonomic studies of leaves are one of the most effective means of correctly identifying plant species. In this paper, mixed activation function is used to improve the ResNet50 network in order to further improve the accuracy of leaf recognition. Firstly, leaf images of 15 common tree species in northern China were collected from the Urban Forestry Demonstration Base of Northeast Forestry University (45°43′–45°44′ N, 126°37′–126°38′ E, forest type was artificial forest), and a small leaf dataset was established. After that, seven commonly used activation functions were selected to improve the ResNet50 network structure, and the improved network was applied to the transfer learning research of the leaf small dataset. On this basis, five activation functions with better performance were selected for the study of mixed activation functions in deep learning. Two of these five activation functions are arbitrarily selected for combination, and a total of twenty combinations are obtained. Further, the first activation function was used in each combination to replace the first ReLU function after all addition operations in the ResNet50 network residual block structure, and another activation function was used to replace the other position ReLU functions. The experimental results show that in the transfer learning of the leaf small dataset using the ResNet50 deep residual network, the appropriate combination of mixed activation functions can increase the performance of the improved network to a certain extent. Among them, the ELU-Swish1 combination has the most significant improvement effect on the network performance, whose final effective validation accuracy reaches 98.17%. Furthermore, the comparison with GoogLeNet and VGG-16 also demonstrates the excellent performance of the improved ELU-Swish1 ResNet50 (ES-ResNet50) network architecture. Finally, tests on the other two small leaf datasets, Flavia and Swedish, also demonstrate the performance improvement of ES-ResNet50. The validation accuracy of the improved ES-Resnet 50 algorithm on these two datasets reaches 99.30% and 99.39%, respectively. All these experiments prove that the recognition performance of leaf transfer learning using the ES-ResNet50 network is indeed improved, which may be caused by the complementarity of the e-exponential gradient of ELU and Swish1 activation functions in the negative region.

In the absence of high-quality hardwood timber resources, we have gradually turned our attention from natural forests to planted fast-growing forests. However, fast-growing tree timber in general has defects such as low wood density, loose texture, and poor mechanical properties. Therefore, improving the performance of wood through efficient and rapid technological processes and increasing the utilization of inferior wood is a good way to extend the use of wood. Densification of wood increases the strength of low-density wood and extends the range of applications for wood and wood-derived products. In this paper, the effects of ultrasonic and vacuum pretreatment on the properties of high-performance wood were explored by combining sonication, vacuum impregnation, chemical softening, and thermomechanical treatments to densify the wood; then, the changes in the chemical composition, microstructure, and mechanical properties of poplar wood before and after treatment were analyzed comparatively by FT-IR, XRD, SEM, and mechanical tests. The results showed that with ultrasonic pretreatment and vacuum impregnation, the compression ratio of high-performance wood reached its highest level and the MOR and MOE reached their maximums. With the help of this method, fast-growing softwoods can be easily prepared into dense wood materials, and it is hoped that this new material can be applied in the fields of construction, aviation, and automobile manufacturing.

In the modern forestry, the demand for renewable and environmentally friendly wood protection is increasing. This paper reports a green method for preparing stable and self-cleaning superhydrophobic coating for wood protection by dripping polyvinyl alcohol cross-linked hollow silica nanoparticles on the surface of wood in combination with polydimethylsiloxane modification. The coating is based on a laminated structure with layers stacked on the surface of the wood and cured quickly with the assistance of UV. The coatings obtained on wood substrates with appropriate ratios have excellent superhydrophobic properties, with an optimum water contact angle of up to 160.4 ± 0.2°. The coating also exhibits good transparency in the UV–visible spectrum and a maximum transmittance of 91%. With transmission electron microscopy, the microscopic morphology of the self-assembled hollow silica nanoparticles was observed. Scanning electron microscopy, Fourier transform infrared spectroscopy, and X-ray diffraction were also applied to investigate the morphology and chemical composition of the coatings. A water contact angle of 151.5 ± 0.7° was maintained even after the abrasion tests with sandpaper at a distance of 300 cm. Meanwhile, the resultant coatings exhibit good self-cleaning properties apart from mechanical durability and chemical stability, which enables effective resistance to contamination. Evidenced by the abovementioned data, this composite coating is capable of optimizing the surface wettability of wood, offering a new dimension to the extensive and prolonged application of wood and wood-based products. Furthermore, considering the advantages of this method, it could also be used in other areas in the future, such as glass, solar substrates, and optical devices.

Effective development and utilization of wood resources is critical. Wood modification research has become an integral dimension of wood science research, however, the similarities between modified wood and original wood render it challenging for accurate identification and classification using conventional image classification techniques. So, the development of efficient and accurate wood classification techniques is inevitable. This paper presents a one-dimensional, convolutional neural network (i.e., BACNN) that combines near-infrared spectroscopy and deep learning techniques to classify poplar, tung, and balsa woods, and PVA, nano-silica-sol and PVA-nano silica sol modified woods of poplar. The results show that BACNN achieves an accuracy of 99.3% on the test set, higher than the 52.9% of the BP neural network and 98.7% of Support Vector Machine compared with traditional machine learning methods and deep learning based methods; it is also higher than the 97.6% of LeNet, 98.7% of AlexNet and 99.1% of VGGNet-11. Therefore, the classification method proposed offers potential applications in wood classification, especially with homogeneous modified wood, and it also provides a basis for subsequent wood properties studies.

In order to improve the mechanical properties of the wood surface and explore the mechanical effect of wood veneer surface metallization, the 31-year-old Pinus sylvestris is taken as the research object and Cu is deposited on the wood surface by magnetron sputtering to achieve wood veneer metallization. Based on X-ray diffraction (XRD) and nanoindentation, a research on the gradual process of the structures and mechanical properties of wood veneer metallization was carried out. The results indicate that wood veneer metallization does not affect the crystallization zone of wood, there are still wood cellulose characteristic peaks and the crystalline structure of the wood cellulose is not damaged; the thickness of the copper thin film increases with the increase of the deposition time, the cellulose characteristic peak strength gradually decrease, and the relative crystallinity also decreases; the characteristic diffraction peaks of Cu (111), Cu (200), and Cu (220) appear near the diffraction angle 2θ which is equal to 43.3°, 50.4°, and 74.1°, and the diffraction peak intensity increases with increase of deposition time, the copper film of the metal wood veneer crystallizes well; the load–displacement of wood veneer decreases significantly with the increase of deposition time, while the moduli of elasticity and hardness increase rapidly. The load–displacement of the samples which were coated for 15 min decreased by 80%, while the moduli of elasticity and hardness of these samples increased by 24.1 times and 17.3 times, respectively. From the results of Scanning Electron Microscope (SEM) measurement of the metallization of wood veneer, it can be seen that the uniform and continuous copper film can be formed on the wood veneer surface by using the magnetron sputtering method. This paper provides a basis for wood veneer surface metallization, which is of great significance for the functional improvement of wood, the expansion of wood application fields, and the enhancement of added value.