Due to the various shapes, textures, and colors of fires, forest fire detection is a challenging task. The traditional image processing method relies heavily on manmade features, which is not universally applicable to all forest scenarios. In order to solve this problem, the deep learning technology is applied to learn and extract features of forest fires adaptively. However, the limited learning and perception ability of individual learners is not sufficient to make them perform well in complex tasks. Furthermore, learners tend to focus too much on local information, namely ground truth, but ignore global information, which may lead to false positives. In this paper, a novel ensemble learning method is proposed to detect forest fires in different scenarios. Firstly, two individual learners Yolov5 and EfficientDet are integrated to accomplish fire detection process. Secondly, another individual learner EfficientNet is responsible for learning global information to avoid false positives. Finally, detection results are made based on the decisions of three learners. Experiments on our dataset show that the proposed method improves detection performance by 2.5% to 10.9%, and decreases false positives by 51.3%, without any extra latency.

Estimating forest structural attributes of planted forests plays a key role in managing forest resources, monitoring carbon stocks, and mitigating climate change. High-resolution and low-cost remote-sensing data are increasingly available to measure three-dimensional (3D) canopy structure and model forest structural attributes. In this study, we compared two suites of point cloud metrics and the accuracies of predictive models of forest structural attributes using unmanned aerial vehicle (UAV) light detection and ranging (LiDAR) and digital aerial photogrammetry (DAP) data, in a subtropical coastal planted forest of East China. A comparison between UAV-LiDAR and UAV-DAP metrics was performed across plots among different tree species, heights, and stem densities. The results showed that a higher similarity between the UAV-LiDAR and UAV-DAP metrics appeared in the dawn redwood plots with greater height and lower stem density. The comparison between the UAV-LiDAR and DAP metrics showed that the metrics of the upper percentiles (r for dawn redwood = 0.95–0.96, poplar = 0.94–0.95) showed a stronger correlation than the lower percentiles (r = 0.92–0.93, 0.90–0.92), whereas the metrics of upper canopy return density (r = 0.21–0.24, 0.14–0.15) showed a weaker correlation than those of lower canopy return density (r = 0.32–0.68, 0.31–0.52). The Weibull α parameter indicated a higher correlation (r = 0.70–0.72) than that of the Weibull β parameter (r = 0.07–0.60) for both dawn redwood and poplar plots. The accuracies of UAV-LiDAR (adjusted (Adj)R² = 0.58–0.91, relative root-mean-square error (rRMSE) = 9.03%–24.29%) predicted forest structural attributes were higher than UAV-DAP (Adj-R² = 0.52–0.83, rRMSE = 12.20%–25.84%). In addition, by comparing the forest structural attributes between UAV-LiDAR and UAV-DAP predictive models, the greatest difference was found for volume (ΔAdj-R² = 0.09, ΔrRMSE = 4.20%), whereas the lowest difference was for basal area (ΔAdj-R² = 0.03, ΔrRMSE = 0.86%). This study proved that the UAV-DAP data are useful and comparable to LiDAR for forest inventory and sustainable forest management in planted forests, by providing accurate estimations of forest structural attributes.

Accurate classification of tree-species is essential for sustainably managing forest resources and effectively monitoring species diversity. In this study, we used simultaneously acquired hyperspectral and LiDAR data from LiCHy (Hyperspectral, LiDAR and CCD) airborne system to classify tree-species in subtropical forests of southeast China. First, each individual tree crown was extracted using the LiDAR data by a point cloud segmentation algorithm (PCS) and the sunlit portion of each crown was selected using the hyperspectral data. Second, different suites of hyperspectral and LiDAR metrics were extracted and selected by the indices of Principal Component Analysis (PCA) and the mean decrease in Gini index (MDG) from Random Forest (RF). Finally, both hyperspectral metrics (based on whole crown and sunlit crown) and LiDAR metrics were assessed and used as inputs to Random Forest classifier to discriminate five tree-species at two levels of classification. The results showed that the tree delineation approach (point cloud segmentation algorithm) was suitable for detecting individual tree in this study (overall accuracy = 82.9%). The classification approach provided a relatively high accuracy (overall accuracy > 85.4%) for classifying five tree-species in the study site. The classification using both hyperspectral and LiDAR metrics resulted in higher accuracies than only hyperspectral metrics (the improvement of overall accuracies = 0.4–5.6%). In addition, compared with the classification using whole crown metrics (overall accuracies = 85.4–89.3%), using sunlit crown metrics (overall accuracies = 87.1–91.5%) improved the overall accuracies of 2.3%. The results also suggested that fewer of the most important metrics can be used to classify tree-species effectively (overall accuracies = 85.8–91.0%).

Canopy cover is a key forest structural parameter that is commonly used in forest inventory, sustainable forest management and maintaining ecosystem services. Recently, much attention has been paid to the use of unmanned aerial vehicle (UAV)-based light detection and ranging (LiDAR) due to the flexibility, convenience, and high point density advantages of this method. In this study, we used UAV-based LiDAR data with individual tree segmentation-based method (ITSM), canopy height model-based method (CHMM), and a statistical model method (SMM) with LiDAR metrics to estimate the canopy cover of a pure ginkgo (Ginkgo biloba L.) planted forest in China. First, each individual tree within the plot was segmented using watershed, polynomial fitting, individual tree crown segmentation (ITCS) and point cloud segmentation (PCS) algorithms, and the canopy cover was calculated using the segmented individual tree crown (ITSM). Second, the CHM-based method, which was based on the CHM height threshold, was used to estimate the canopy cover in each plot. Third, the canopy cover was estimated using the multiple linear regression (MLR) model and assessed by leave-one-out cross validation. Finally, the performance of three canopy cover estimation methods was evaluated and compared by the canopy cover from the field data. The results demonstrated that, the PCS algorithm had the highest accuracy (F = 0.83), followed by the ITCS (F = 0.82) and watershed (F = 0.79) algorithms; the polynomial fitting algorithm had the lowest accuracy (F = 0.77). In the sensitivity analysis, the three CHM-based algorithms (i.e., watershed, polynomial fitting and ITCS) had the highest accuracy when the CHM resolution was 0.5 m, and the PCS algorithm had the highest accuracy when the distance threshold was 2 m. In addition, the ITSM had the highest accuracy in estimation of canopy cover (R² = 0.92, rRMSE = 3.5%), followed by the CHMM (R² = 0.94, rRMSE = 5.4%), and the SMM had a relative low accuracy (R² = 0.80, rRMSE = 5.9%).The UAV-based LiDAR data can be effectively used in individual tree crown segmentation and canopy cover estimation at plot-level, and CC estimation methods can provide references for forest inventory, sustainable management and ecosystem assessment.

Accurate and timely estimation of forest structural parameters plays a key role in the management of forest resources, as well as studies on the carbon cycle and biodiversity. Light Detection and Ranging (LiDAR) is a promising active remote sensing technology capable of providing highly accurate three dimensional and wall-to-wall forest structural characteristics. In this study, we evaluated the utility of standard metrics and canopy metrics derived from airborne LiDAR data for estimating plot-level forest structural parameters individually and in combination, over a subtropical forest in Yushan forest farm, southeastern China. Standard metrics, i.e., height-based and density-based metrics, and canopy metrics extracted from canopy vertical profiles, i.e., canopy volume profile (CVP), canopy height distribution (CHD), and foliage profile (FP), were extracted from LiDAR point clouds. Then the standard metrics and canopy metrics were used for estimating forest structural parameters individually and in combination by multiple regression models, including forest type-specific (coniferous forest, broad-leaved forest, mixed forest) models and general models. Additionally, the synergy of standard metrics and canopy metrics for estimating structural parameters was evaluated using field measured data. Finally, the sensitivity of vertical and horizontal resolution of voxel size for estimating forest structural parameters was assessed. The results showed that, in general, the accuracies of forest type-specific models (Adj-R² = 0.44–0.88) were relatively higher than general models (Adj-R² = 0.39–0.77). For forest structural parameters, the estimation accuracies of Lorey’s mean height (Adj-R² = 0.61–0.88) and aboveground biomass (Adj-R² = 0.54–0.81) models were the highest, followed by volume (Adj-R² = 0.42–0.78), DBH (Adj-R² = 0.48–0.74), basal area (Adj-R² = 0.41–0.69), whereas stem density (Adj-R² = 0.39–0.64) models were relatively lower. The combination models (Adj-R² = 0.45–0.88) had higher performance compared with models developed using standard metrics (only) (Adj-R² = 0.42–0.84) and canopy metrics (only) (Adj-R² = 0.39–0.83). The results also demonstrated that the optimal voxel size was 5 × 5 × 0.5 m³ for estimating most of the parameters. This study demonstrated that canopy metrics based on canopy vertical profiles can be effectively used to enhance the estimation accuracies of forest structural parameters in subtropical forests.

Rubber trees in southern China are often impacted by natural disturbances that can result in a tilted tree body. Accurate crown segmentation for individual rubber trees from scanned point clouds is an essential prerequisite for accurate tree parameter retrieval. In this paper, three plots of different rubber tree clones, PR107, CATAS 7-20-59, and CATAS 8-7-9, were taken as the study subjects. Through data collection using ground-based mobile light detection and ranging (LiDAR), a voxelisation method based on the scanned tree trunk data was proposed, and deep images (i.e., images normally used for deep learning) were generated through frontal and lateral projection transform of point clouds in each voxel with a length of 8 m and a width of 3 m. These images provided the training and testing samples for the faster region-based convolutional neural network (Faster R-CNN) of deep learning. Consequently, the Faster R-CNN combined with the generated training samples comprising 802 deep images with pre-marked trunk locations was trained to automatically recognize the trunk locations in the testing samples, which comprised 359 deep images. Finally, the point clouds for the lower parts of each trunk were extracted through back-projection transform from the recognized trunk locations in the testing samples and used as the seed points for the region’s growing algorithm to accomplish individual rubber tree crown segmentation. Compared with the visual inspection results, the recognition rate of our method reached 100% for the deep images of the testing samples when the images contained one or two trunks or the trunk information was slightly occluded by leaves. For the complicated cases, i.e., multiple trunks or overlapping trunks in one deep image or a trunk appearing in two adjacent deep images, the recognition accuracy of our method was greater than 90%. Our work represents a new method that combines a deep learning framework with point cloud processing for individual rubber tree crown segmentation based on ground-based mobile LiDAR scanned data.