Tree Species Recognition Research Articles

DETERMINING THE SPECIES COMPOSITION OF FOREST VEGETATION IN THE KOSTANAY REGION USING REMOTE SENSING DATA

During the scientific investigation, woody species of forest vegetation were identified and a map of forest vegetation in the Kostanay region was produced using various data sources: field materials, Earth remote sensing data, and ArcGIS10.9 software. An algorithm was developed to detect tree species based on Landsat 9 satellite imagery, characterized by high spatial resolution. Recognition of dominant tree species was performed using various combinations of spectral bands from Landsat 9 imagery, analysis of vegetation indices (NDVI, EVI) across different seasons, and supervised local adaptive classification. The obtained data were validated against field research materials (August-September 2023) and forest management records. The chosen algorithm implements contemporary approaches to acquiring and processing necessary data from satellite remote sensing imagery. Further differentiation and creation of the forest vegetation map of the Kostanay region were based on the established map of tree species, digital elevation model, geological-geomorphological features, field research, thematic maps, and physical geography of the region. As a result of the conducted research, six classes of forest stands were delineated in the Kostanay region, including light-coniferous and deciduous tree species such as pine, birch, aspen, larch, shrubbery, and meadow vegetation. These research findings and the algorithm developed can be applied to other study areas and hold practical significance.

A Hybrid convolution neural network for the classification of tree species using hyperspectral imagery.

In recent years, the advancement of hyperspectral remote sensing technology has greatly enhanced the detailed mapping of tree species. Nevertheless, delving deep into the significance of hyperspectral remote sensing data features for tree species recognition remains a challenging endeavor. The method of Hybrid-CS was proposed to addresses this challenge by synergizing the strengths of both deep learning and traditional learning techniques. Initially, we extract comprehensive correlation structures and spectral features. Subsequently, a hybrid approach, combining correlation-based feature selection with an optimized recursive feature elimination algorithm, identifies the most valuable feature set. We leverage the Support Vector Machine algorithm to evaluate feature importance and perform classification. Through rigorous experimentation, we evaluate the robustness of hyperspectral image-derived features and compare our method with other state-of-the-art classification methods. The results demonstrate: (1) Superior classification accuracy compared to traditional machine learning methods (e.g., SVM, RF) and advanced deep learning approaches on the tree species dataset. (2) Enhanced classification accuracy achieved by incorporating SVM and CNN information, particularly with the integration of attention mechanisms into the network architecture. Additionally, the classification performance of a two-branch network surpasses that of a single-branch network. (3) Consistent high accuracy across different proportions of training samples, indicating the stability and robustness of the method. This study underscores the potential of hyperspectral images and our proposed methodology for achieving precise tree species classification, thus holding significant promise for applications in forest resource management and monitoring.

Tree species classification on images from airborne mobile mapping using ML.NET

ABSTRACT Deep learning is a powerful tool for automating the process of recognizing and classifying objects in images. In this study, we used ML.NET, a popular open-source machine learning framework, to develop a model for identifying tree species in images obtained from airborne mobile mapping. These high-resolution images can be used to create detailed maps of the landscape. They can also be analyzed and processed to extract information about visual features, including tree species recognition. The deep learning model was trained using ML.NET to classify two tree species based on the combination of airborne mobile mapping images. Our approach yielded impressive results, with a maximum classification accuracy of 93.9%. This demonstrates the effectiveness of combining imagery sources with deep learning tools in ML.NET for efficient and accurate tree species classification. This study highlights the potential of the ML.NET framework for automating object classification and can provide valuable insights and information for forestry management and conservation efforts. The primary objective of this research was to evaluate the effectiveness of an approach for identifying tree species through a model generated using a combination of ortho and oblique images captured by a mobile mapping system.

HPAC: a forest tree species recognition network based on multi- scale spatial enhancement in remote sensing images

ABSTRACT Forest tree species recognition is a pivotal subject in the field of remote sensing. To address this, deep learning has been extensively applied. Thus far, most classification methods have generally relied on learning certain global features, yet often overlook the characteristics of specific regions, thereby struggling to adequately handle the similarity between classes. Furthermore, due to the singular nature of features, effectively representing the attributes of tree species images becomes challenging, consequently impacting classification performance. To tackle these issues, a novel approach for forest tree species classification in remote sensing images is proposed, based on the Hollow Pyramid Attention Combination (HPAC) network. Initially, a Shallow Multi-scale Hollow Fusion (SMHF) module is introduced before the 7 × 7 convolution in the ResNet-50 network and the first residual block’s first layer. This module employs dilated convolutions to achieve varying receptive fields. Moreover, it incorporates positional feature information, significantly enhancing the shallow-level feature extraction capabilities, resulting in a richer feature representation. Subsequently, to minimize network parameters and computational workload while bolstering the capacity to recognize deep-level features, the last residual block of the ResNet-50 differentiation is substituted with a Maxpool Avgpool Fusion (MAF) module. This replacement serves to enhance classification accuracy. The classification process is ultimately concluded with a Softmax classifier. Experimental results underscore the effectiveness of the proposed method, achieving a classification accuracy of 95.89% on the PCANDVI dataset of forest tree species data (FTSD). In summary, the introduced HPAC network proves to be both feasible and effective.

A Multi-Scale Convolution and Multi-Layer Fusion Network for Remote Sensing Forest Tree Species Recognition

Forest tree species identification in the field of remote sensing has become an important research topic. Currently, few research methods combine global and local features, making it challenging to accurately handle the similarity between different categories. Moreover, using a single deep layer for feature extraction overlooks the unique feature information at intermediate levels. This paper proposes a remote sensing image forest tree species classification method based on the Multi-Scale Convolution and Multi-Level Fusion Network (MCMFN) architecture. In the MCMFN network, the Shallow Multi-Scale Convolution Attention Combination (SMCAC) module replaces the original 7 × 7 convolution at the first layer of ResNet-50. This module uses multi-scale convolution to capture different receptive fields, and combines it with the attention mechanism to effectively enhance the ability of shallow features and obtain richer feature information. Additionally, to make efficient use of intermediate and deep-level feature information, the Multi-layer Selection Feature Fusion (MSFF) module is employed to improve classification accuracy. Experimental results on the Aerial forest dataset demonstrate a classification accuracy of 91.03%. The comprehensive experiments indicate the feasibility and effectiveness of the proposed MCMFN network.

Spectral volume index creation and performance evaluation: A preliminary test for tree species identification

To fully mine information regarding differences among various tree species from remote sensing data and improve the accuracy of tree species recognition, this study utilized the spectral reflection value, wavelength, and time as parameters and employed three algorithms to create an expression for the spectral volume index (SVI). Then, data were obtained by applying RedEdge-MX to four phases, SVI features were extracted, and a mixed feature set of spectral band + texture + digital surface model + SVI was constructed. A random forest algorithm was employed to determine the importance of the SVI features and derive the optimal feature set for tree species classification. The additional objectives were to determine if the SVI features have an active role in tree species classification and which algorithm is more conducive for extracting useful SVI features. The SVI features extracted with volume constraints exhibit better performance in tree species recognition than those extracted without volume constraints. Moreover, the SVI features extracted using a variable-constrained volume were better than those extracted using a constant-constrained volume. The combination of SVI features could improve the accuracy of tree species recognition (the highest overall accuracy was 92.76%), but the improvement effect was limited (the value was 92.16% when SVI features were not combined). These findings show that the SVI obtained using this method could be used to mine the difference information of tree species in images to a certain extent and hence, could be used in tree species identification.

Leaves and Twigs Image Recognition Based on Deep Learning and Combined Classifier Algorithms

In recent years, the automatic recognition of tree species based on images taken by digital cameras has been widely applied. However, many problems still exist, such as insufficient tree species image acquisition, uneven distribution of image categories, and low recognition accuracy. Tree leaves can be used to differentiate and classify tree species due to their cognitive signatures in color, vein texture, shape contour, and edge serration. Moreover, the way the leaves are arranged on the twigs has strong characteristics. In this study, we first built an image dataset of 21 tree species based on the features of the twigs and leaves. The tree species feature dataset was divided into the training set and test set, with a ratio of 8:2. Feature extraction was performed after training the convolutional neural network (CNN) using the k-fold cross-validation (K-Fold–CV) method, and tree species classification was performed with classifiers. To improve the accuracy of tree species identification, we combined three improved CNN models with three classifiers. Evaluation indicators show that the overall accuracy of the designed composite model was 1.76% to 9.57% higher than other CNN models. Furthermore, in the MixNet XL CNN model, combined with the K-nearest neighbors (KNN) classifier, the highest overall accuracy rate was obtained at 99.86%. In the experiment, the Grad-CAM heatmap was used to analyze the distribution of feature regions that play a key role in classification decisions. Observation of the Grad-CAM heatmap illustrated that the main observation area of SE-ResNet50 was the most accurately positioned, and was mainly concentrated in the interior of small twigs and leaflets. Our research showed that modifying the training method and classification module of the CNN model and combining it with traditional classifiers to form a composite model can effectively improve the accuracy of tree species recognition.

The image recognition of urban greening tree species based on deep learning and CAMP-MKNet model

The information of urban tree species resources is of vital significance to the planning and design of urban green spaces. Tree organs, such as the bark are used as the primary features of identifying tree species. However, traditional tree identification methods need to consume a lot of manpower and time costs. In addition, the application of machine image recognition technology to tree species recognition still has problems such as heavy data preprocessing workload, small number of tree species images, uneven distribution of categories, and low recognition accuracy. In order to promote the intelligent management of urban forestry and solve the above problems, it is necessary to establish an automatic image recognition model for urban greening tree species. We captured bark images of 21 urban afforestation tree species in their natural environment and constructed a dataset that was divided into a train set, validation set, and test set in the ratio of 7:1:2. Combining Channel Attention Module (CAM) with algorithms such as Spatial Pyramid Pooling (SPP) and Mixed Depthwise Dilated Convolutional Kernels. The core algorithm is Mixed Convolution Kernel (MK), and a CAMP-MKNet Convolutional Neural Network (CNN) is constructed as a bark image classification model for urban greening tree species. The overall accuracy of the generic models ranged from 41.06% to 82.03%, whereas the overall accuracy of the experimental CAMP-MKNet model was 84.25%, with lower prediction cost. Our study shows that the CAMP-MKNet CNN model with better prediction performance and computational cost and can provide crucial insights and technical support for developing automated urban tree species image recognition systems.

Multispectral versus texture features from ZiYuan-3 for recognizing on deciduous tree species with cloud and SVM models

Tree species recognition accuracy greatly affects forest remote sensing mapping and forestry resource monitoring. The multispectral and texture features of the remote sensing images from the ZiYuan-3 (ZY-3) satellite at two phenological phases of autumn and winter (September 29th and December 7th) were selected for constructing and optimizing sensitive spectral indices and texture indices. Multidimensional cloud model and support vector machine (SVM) model were constructed by the screened spectral and texture indices for remote sensing recognition of Quercus acutissima (Q. acutissima) and Robinia pseudoacacia (R. pseudoacacia) on Mount Tai. The results showed that, the correlation intensities of the constructed spectral indices with tree species were preferable in winter than in autumn. The spectral indices constructed by band 4 showed the superior correlation compared with other bands, both in the autumn and winter time phases. The optimal sensitive texture indices for both phases were mean, homogeneity and contrast for Q. acutissima, and contrast, dissimilarity and second moment for R. pseudoacacia. Spectral features were found to have a higher recognition accuracy than textural features for recognizing on both Q. acutissima and R. pseudoacacia, and winter showing superior recognition accuracy than autumn, especially for Q. acutissima. The recognition accuracy of the multidimensional cloud model (89.98%) does not show a superior advantage over the one-dimensional cloud model (90.57%). The highest recognition accuracy derived from a three-dimensional SVM was 84.86%, which was lower than the cloud model (89.98%) in the same dimension. This study is expected to provide technical support for the precise recognition and forestry management on Mount Tai.

Effectiveness of the spectral area index created by three algorithms for tree species recognition

Key messageTree species identification analysis of the two images (Luoyang and Hohhot of China) shows that the polygonal area indices extracted by the specific band-constrained polygon relative area (algorithm 3, obtained accuracy was ~ 13% higher than that of other algorithms in WorldView-3 and ~ 2% higher in WorldView-2) can effectively improve the classification accuracy of tree species compared to those with a constant polygon relative area constraint (algorithm 2) and without area constraint (algorithm 1) (equal accuracy was obtained by algorithms 1 and 2 in each data).ContextSolving the problem of tree species identification by remote sensing technology is an international issue. Exploring the improvement of tree species recognition accuracy through multiple methods is currently widely attempted. A previous study has indicated that mining the differential information of various tree species in images using area differences of the polygons formed by tree species spectral curves and creating the polygon area index can improve tree species recognition accuracy. However, this study only created two such indices. Thus, a general model was developed to extract more potential polygon area indices and help tree species classification. However, the improvement of this model using a constant and a specific band to constrain the relative area of polygons still needs to be fully studied.AimsTo obtain new algorithms for extracting polygon area indices that can mine the differential information of tree species and determine the index that is the most effective for tree species classification.MethodsBy unconstraining the area of polygons and constraining the relative area of polygons with constant and specific bands, three formulations of polygon area indices were created. Polygon area indices were extracted from WorldView-3 and WorldView-2 imagery based on three algorithms and combined with textures and spectral bands to form three feature sets. Random forest was used to classify images and rank the importance of features in the feature sets, and the effectiveness of the polygon area indices extracted by each algorithm in tree species recognition was analysed in accordance with their performance in the classifications.ResultsThe proportion of polygon area index in the optimal feature sets ranged from 36.4 to 63.1%. The polygon area indices extracted with constant constrained polygon relative area and those without area constraint have minimal effect on tree species classification accuracy. Meanwhile, the polygon area indices extracted by the algorithm of specific band-constrained polygon relative area could remarkably improve tree species recognition accuracy (compared with spectral bands, WorldView-3 and WorldView-2 improved by 9.69% and 4.19%, respectively).ConclusionThe experiments confirmed that polygon area indices are beneficial for tree species classification, and polygon area indices extracted by specific band-constrained polygon relative area play an important role in tree species identification.

AMDNet: A Modern UAV RGB Remote-Sensing Tree Species Image Segmentation Model Based on Dual-Attention Residual and Structure Re-Parameterization

Tree species classification is an important and challenging task in image recognition and the management of forest resources. Moreover, the task of tree species classification based on remote sensing images can significantly improve the efficiency of the tree species survey and save costs. In recent years, many large models have achieved high accuracy in the task of tree species classification in an airborne remote-sensing manner, but due to their fixed geometric structure, traditional convolutional neural networks are inherently limited to the local receptive field and can only provide segmental context information. The limitation of insufficient context information greatly affects the segmentation accuracy. In this paper, a dual-attention residual network (AMDNet) and a re-parameterized model approach are proposed to capture the global context information, fuse the weight, reduce the model volume, and maintain the computational efficiency. Firstly, we propose MobileNetV2 as the backbone network for feature extraction to further improve the feature identification by modeling semantic dependencies in the spatial dimension and channel dimension and adding the output of the two attention modules. Then, the attention perception features are generated by stacking the attention modules, and the in-depth residual attention network is trained using attention residual learning, through which more accurate segmentation results can be obtained. Secondly, we adopt the approach of structure re-parameterization, use a multi-branch topology for training, carry out weighted averaging on multiple trained models, and fuse multiple branch modules into a completely equivalent module in inference. The proposed approach results in a reduction in the number of parameters and an accelerated inference speed while also achieving improved classification accuracy. In addition, the model training strategy is optimized based on Transformer to enhance the accuracy of segmentation. The model was used to conduct classification experiments on aerial orthophotos of Hongya Forest Farm in Sichuan, China, and the MIOU of tree species recognition using the test equipment reached 93.8%. Compared with current models such as UNet, our model exhibits a better performance in terms of both speed and accuracy, in addition to its enhanced deployment capacity, and its speed advantage is more conducive to real-time segmentation, thereby representing a novel approach for the classification of tree species in remote sensing imagery with significant potential for practical applications.

Tree Species Recognition Using Combined Attention and ResNet for Unmanned Aerial Vehicle Images

Tree Species Recognition Using Combined Attention and ResNet for Unmanned Aerial Vehicle Images

Classification of tree species based on hyperspectral reflectance images of stem bark

ABSTRACT Automatization of tree species identification in the field is crucial in improving forest-based bioeconomy, supporting forest management, and facilitating in situ data collection for remote sensing applications. However, tree species recognition has never been addressed with hyperspectral reflectance images of stem bark before. We investigated how stem bark texture differs between tree species using a hyperspectral camera set-up and gray level co-occurrence matrices and assessed the potential of using reflectance spectra and texture features of stem bark to identify tree species. The analyses were based on 200 hyperspectral reflectance data cubes (415–925 nm) representing ten tree species. There were subtle interspecific differences in bark texture. Using average spectral features in linear discriminant analysis classifier resulted in classification accuracy of 92–96.5%. Using spectral and texture features together resulted in accuracy of 93–97.5%. With a convolutional neural network, we obtained an accuracy of 94%. Our study showed that the spectral features of stem bark were robust for classifying tree species, but importantly, bark texture is beneficial when combined with spectral data. Our results suggest that in situ imaging spectroscopy is a promising sensor technology for developing accurate tree species identification applications to support remote sensing.

Remote Sensing Extraction Method of Illicium verum Based on Functional Characteristics of Vegetation Canopy

With the rapid development of remote sensing technology, researchers have attempted to improve the accuracy of tree species classifications from both data sources and methods. Although previous studies on tree species recognition have utilized the spectral and textural features of remote sensing images, they are unable to effectively extract tree species due to the problems of “same object with different spectrum” and “foreign object with the same spectrum”. Therefore, this study introduces vegetation functional datasets to further improve tree species classification. Using vegetation functional datasets, Sentinel-2 (S2) spectral datasets, and environmental datasets, combined with a Random Forest (RF) model, the classification of six types of land cover in Leye, Guangxi was completed and the planting distribution of Illicium verum in Leye County was extracted. Our results showed that the combination of vegetation functional datasets, S2 spectral datasets, and environmental datasets provided the highest overall accuracy (OA) (0.8671), Kappa coefficient (0.8382), and F1-Score (0.79). We believe that the vegetation functional datasets can enhance the accuracy of Illicium verum classification and provide new directions for tree species identification research. If vegetation functional datasets from more tree species are obtained in the future, we can extend them to the level of multiple tree species, and this approach may help to extract more information about forest species from remote sensing data in future studies.

Tree Species Classification Using Plant Functional Traits and Leaf Spectral Properties along the Vertical Canopy Position

Plant functional traits are rarely used in tree species classification, and the impact of vertical canopy positions on collecting samples for classification also remains unclear. We aim to explore the feasibility and effectiveness of leaf traits in classification, as well as to detect the effect of vertical position on classification accuracy. This work will deepen our understanding of the ecological mechanism of natural forest structure and succession from new perspectives. In this study, we collected foliar samples from three canopy layers (upper, middle and lower) and measured their spectra, as well as eight well-known leaf traits. We used a leaf hyperspectral reflectance (LHR) dataset, leaf functional traits (LFT) dataset and LFT + LHR dataset to classify six dominant tree species in a subtropical evergreen broad-leaved forest. Our results showed that the LFT + LHR dataset achieved the highest classification results (overall accuracy (OA) = 77.65% and Kappa = 0.73), followed by the LFT dataset (OA = 74.26% and Kappa = 0.69) and the LHR dataset (OA = 69.06% and Kappa = 0.63). Along the vertical canopy, the OA and Kappa increased from the lower to the upper layers, and the combination data of the three canopy layers achieved the highest accuracy. For the individual tree species, the shade-tolerant species (including Machilus chinensis, Cryptocarya chinensis and Cryptocarya concinna) produced higher accuracies than the light-demanding species (including Schima superba and Castanopsis chinensis). Our results provide an approach for enhancing tree species recognition from the plant physiology and biochemistry perspective and emphasize the importance of vertical direction in forest community research.

Establishing a citywide street tree inventory with street view images and computer vision techniques

Trees in urban areas have diverse ecological, social, and health benefits. The establishment of up-to-date and accurate street-tree inventories that list the species and locations of individual street trees is critical to urban tree management and tree-planting campaigns. However, street-tree inventories are incomplete or lacking altogether in most cities. This is partly because conventional field assessment is laborious or expensive. In this study, we developed and validated a novel and cost-effective method to establish a city-wide tree inventory based on computer vision and freely available street view images (SVIs). Tree information such as species, height, crown diameter, and geographical coordinates at the individual tree level can be assessed. Based on an object detection model, we adopted a species-based loss function to address the challenges of long-tailed class distribution of species, which is caused by imbalance among sample size of different tree species and can lead to poor performance of the model. Compared with other research in urban tree species recognition, the modified model shows a higher accuracy. In order to calculate quantitative features of street trees, we employed a deep learning algorithm, which is pretrained on stereo dataset and validated on Google Street View images, to estimate the depth of each pixel in SVIs. Furthermore, as a demonstration, we established the citywide tree inventory and conducted tree diversity analysis for Jinan, China. Compared with new developed area, the old town has more street trees and more diverse tree species which can improve biodiversity and walkability. We also found that plane trees, which can cause allergic reactions, are dominant in northern new developed urban area.

Optimization Method of Airborne LiDAR Individual Tree Segmentation Based on Gaussian Mixture Model

Forests are the main part of the terrestrial ecosystem. Airborne LiDAR is fast, comprehensive, penetrating, and contactless and can depict 3D canopy information with a high efficiency and accuracy. Therefore, it plays an important role in forest ecological protection, tree species recognition, carbon sink calculation, etc. Accurate recognition of individual trees in forests is a key step to various application. In real practice, however, the accuracy of individual tree segmentation (ITS) is often compromised by under-segmentation due to the diverse species, obstruction and understory trees typical of a high-density multistoried mixed forest area. Therefore, this paper proposes an ITS optimization method based on Gaussian mixture model for airborne LiDAR data. First, the mean shift (MS) algorithm is used for the initial ITS of the pre-processed airborne LiDAR data. Next, under-segmented samples are extracted by integrated learning, normally segmented samples are classified by morphological approximation, and the approximate distribution uncertainty of the normal samples is described with a covariance matrix. Finally, the class composition among the under-segmented samples is determined, and the under-segmented samples are re-segmented using Gaussian mixture model (GMM) clustering, in light of the optimal covariance matrix of the corresponding categories. Experiments with two datasets, Trento and Qingdao, resulted in ITS recall of 94% and 96%, accuracy of 82% and 91%, and F-scores of 0.87 and 0.93. Compared with the MS algorithm, our method is more accurate and less likely to under-segment individual trees in many cases. It can provide data support for the management and conservation of high-density multistoried mixed forest areas.

Tree Species Classification Using Ground-Based LiDAR Data by Various Point Cloud Deep Learning Methods

Tree species information is an important factor in forest resource surveys, and light detection and ranging (LiDAR), as a new technical tool for forest resource surveys, can quickly obtain the 3D structural information of trees. In particular, the rapid and accurate classification and identification of tree species information from individual tree point clouds using deep learning methods is a new development direction for LiDAR technology in forest applications. In this study, mobile laser scanning (MLS) data collected in the field are first pre-processed to extract individual tree point clouds. Two downsampling methods, non-uniform grid and farthest point sampling, are combined to process the point cloud data, and the obtained sample data are more conducive to the deep learning model for extracting classification features. Finally, four different types of point cloud deep learning models, including pointwise multi-layer perceptron (MLP) (PointNet, PointNet++, PointMLP), convolution-based (PointConv), graph-based (DGCNN), and attention-based (PCT) models, are used to classify and identify the individual tree point clouds of eight tree species. The results show that the classification accuracy of all models (except for PointNet) exceeded 0.90, where the PointConv model achieved the highest classification accuracy for tree species classification. The streamlined PointMLP model can still achieve high classification accuracy, while the PCT model did not achieve good accuracy in the tree species classification experiment, likely due to the small sample size. We compare the training process and final classification accuracy of the different types of point cloud deep learning models in tree species classification experiments, further demonstrating the advantages of deep learning techniques in tree species recognition and providing experimental reference for related research and technological development.

Landscape tree species recognition using RedEdge-MX: Suitability analysis of two different texture extraction forms under MLC and RF supervision

Abstract The size of the texture extraction window impacts image tree species classification, and the determination of the optimal texture extraction window requires the supervision of a specific classifier for accuracy. Therefore, it is necessary to analyse which kind of classifier is more suitable and should be to choose. In this study, we extracted eight types of textures, namely mean, variance, homogeneity, contrast, dissimilarity, entropy, second moment and correlation, changed the window size by gradient increase and used maximum likelihood classification (MLC) and random forest (RF) to supervise and determine their optimal extraction windows, respectively. Finally, the optimised time consumption and classification accuracy for tree species classification was identified. The time consumption of MLC was significantly less than that of RF; however, neither was very long; for most textures, the optimal texture extraction window determined by MLC supervision was larger than that determined by RF supervision; in the classification of most feature sets, the overall accuracy obtained by MLC was less than that of RF. Because the time consumption of the texture extraction was much greater than that of the image classification, the comprehensive trade-off indicates that using RF supervision to determine the optimal window for texture extraction was more conducive to tree species recognition.

Optimum Phenological Phases for Deciduous Species Recognition: A Case Study on Quercus acutissima and Robinia pseudoacacia in Mount Tai

Tree species recognition is important for remote–sensing mapping and dynamic monitoring of forest resource. However, the complex phenological cycle poses a challenge to remote–sensing recognition of deciduous tree species in mountainous areas, and the selection of temporal phase is particularly important to improve recognition accuracy. Multispectral images of Ziyuan–1 02C (ZY–1 02C) and Ziyuan–3 (ZY–3) at three phenological phases of spring, autumn and winter (12 May, 29 September and 7 December, recorded as T5–12, T9–29 and T12–7) are selected to optimize sensitive spectral indices. Support vector machine (SVM) and maximum likelihood model (MLE) are constructed to explore the optimum phase of recognizing on Quercus acutissima (O. acutissima ) and Robinia pseudoacacia (R. pseudoacacia) in Mount Tai. The results showed the average spectral reflection intensity of O. acutissima was higher than that of R. pseudoacacia Compared to other phenological periods, the most significant spectral differences between O. acutissima and R. pseudoacacia were found in the spring (12 May), which was identified as the optimum phenological phase. Band 4 is the most sensitive band in all the three phases for the tree species recognition. Moreover, the overall recognition accuracy of deciduous tree species on 12 May reached 89.25%, which was significantly higher than the other two phases. On 12 May, the recognition accuracies of SVM based on sensitive spectral indices of up to 93.59% for O. acutissima and 85.44% for R. pseudoacacia, were higher overall than that of the MLE. Sensitive spectral indices introduced were shown to significantly improve the recognition accuracy for tree species over a single sensitive band. The study is expected to facilitate the precise recognition and forestry management on Mount Tai.