The aim of this study was to select the optimal deep learning model for land cover classification through hyperparameter adjustment. A U-Net model with encoder and decoder structures was used as the deep learning model, and RapidEye satellite images and a sub-divided land cover map provided by the Ministry of Environment were used as the training dataset and label images, respectively. According to different combinations of hyperparameters, including the size of the input image, the configuration of convolutional layers, the kernel size, and the number of pooling and up-convolutional layers, 90 deep learning models were built, and the model performance was evaluated through the training accuracy and loss, as well as the validation accuracy and loss values. The evaluation results showed that the accuracy was higher with a smaller image size and a smaller kernel size, and was more dependent on the convolutional layer configuration and number of layers than the kernel size. The loss tended to be lower as the convolutional layer composition and number of layers increased, regardless of the image size or kernel size. The deep learning model with the best performance recorded a validation loss of 0.11 with an image size of 64 × 64, a convolutional layer configuration of C→C→C→P, a kernel size of 5 × 5, and five layers. Regarding the classification accuracy of the land cover map constructed using this model, the overall accuracy and kappa coefficient for three study cities showed high agreement at approximately 82.9% and 66.3%, respectively.

Stem volume is a critical factor in managing and evaluating forest resources. At present, stem volume is commonly estimated indirectly by constructing a taper model that utilizes sampling, diameter at breast height (DBH), and tree height. However, these estimates are constrained by errors arising from spatial and stand environment variations as well as uncertainties in height measurements. To address these issues, this study aimed to accurately estimate stem volume using light detection and ranging (LiDAR) technology, a key tool in modern precision forestry. LiDAR data were used to build comprehensive three-dimensional models of forests with multi-platform LiDAR systems. This approach allowed for precise measurements of tree heights and stem diameters at various heights, effectively mitigating the limitations of earlier measurement methods. Based on these data, a Kozak taper curve was developed at the individual tree level using LiDAR-derived tree height and diameter estimates. Integrating this curve with LiDAR data enabled a hybrid approach to estimating stem volume, facilitating the calculation of diameters at points not directly identifiable from LiDAR data alone. The proposed method was implemented and evaluated for two economically significant tree species in Korea: Pinus koraiensis and Larix kaempferi. The RMSE comparison between the taper curve-based approach and the hybrid volume estimation method showed that, for Pinus koraiensis, the RMSE was 0.11 m³ using the taper curve-based approach and 0.07 m³ for the hybrid method, while for Larix kaempferi, the RMSE was 0.13 m³ and 0.05 m³, respectively. The estimation error of the hybrid method was approximately half that of the taper curve-based approach. Consequently, the hybrid volume estimation method, which integrates LiDAR and the taper model, overcomes the limitations of conventional taper curve-based approaches and contributes to improving the accuracy of forest resource monitoring.

A growing societal interest exists in the application of lidar technology to monitor forest resource information and forestry management activities. This study examined the possibility of estimating the diameter at breast height (DBH) of two tree species, Pinus koraiensis (PK) and Larix kaempferi (LK), by varying the number of terrestrial laser scanning (TLS) scans (1, 3, 5, 7, and 9) and DBH estimation methods (circle fitting [CF], ellipse fitting [EF], circle fitting with RANSAC [RCF], and ellipse fitting with RANSAC [REF]). This study evaluates the combination that yields the highest estimation accuracy. The results showed that for PK, the lowest RMSE of 0.97 was achieved when REF was applied to the data from nine scans after noise removal. For LK, the lowest RMSE of 1.03 was observed when applying CF to the data from seven scans after noise removal. Furthermore, ANOVA revealed no significant difference in the estimated DBH from nine scans when more than three scans were used for CF and RCF and more than five for EF and REF. These results are expected to be useful in establishing efficient and accurate DBH estimation plans using TLS for forest resource monitoring.

In artificial forests, regular thinning is required to promote diameter growth for producing high-quality, large-diameter timber. However, selecting trees for thinning often relies on qualitative and subjective assessments by field workers. The purpose of this study was to develop a quantitative method for selecting trees for thinning by combining machine learning algorithms and competition indices. Our study site included the Pinus koraiensis area within a Kangwon National University research forest in the Republic of Korea. Data from a model development site were used for the basic crown classification model for Pinus koraiensis. The model was optimized by adjusting hyperparameters. Different algorithms, including Random Forest, XGBoost, and LightGBM (LGBM), were improved using Random Search. LGBM showed the highest accuracy of 71.6%. LGBM—in combination with the competition indices—was used to classify the crown class in the application site and select trees for thinning. Compared to the combination of Braathe and Martin-EK indices, the combination of LGBM and Hegyi index enabled the even distribution of the residual stand in the entire site after thinning. It lowered the distribution of hot spots, which represent competition. Thus, the combination of LGBM and Hegyi index was the most effective option to improve the spatial distribution of trees after thinning. Our findings can improve forest management by providing a quantitative and objective method for selecting trees for thinning.