Machine Learning for Tree Species Classification Using Sentinel-2 Spectral Information, Crown Texture, and Environmental Variables

Abstract

The most recent forest-type map of the Korean Peninsula was produced in 1910. That of South Korea alone was produced since 1972; however, the forest type information of North Korea, which is an inaccessible region, is not known due to the separation after the Korean War. In this study, we developed a model to classify the five dominant tree species in North Korea (Korean red pine, Korean pine, Japanese larch, needle fir, and Oak) using satellite data and machine-learning techniques. The model was applied to the Gwangneung Forest area in South Korea; the Mt. Baekdu area of China, which borders North Korea; and to Goseong-gun, at the border of South Korea and North Korea, to evaluate the model’s applicability to North Korea. Eighty-three percent accuracy was achieved in the classification of the Gwangneung Forest area. In classifying forest types in the Mt. Baekdu area and Goseong-gun, even higher accuracies of 91% and 90% were achieved, respectively. These results confirm the model’s regional applicability. To expand the model for application to North Korea, a new model was developed by integrating training data from the three study areas. The integrated model’s classification of forest types in Goseong-gun (South Korea) was relatively accurate (80%); thus, the model was utilized to produce a map of the predicted dominant tree species in Goseong-gun (North Korea).

Exploitation of Time Series Sentinel-2 Data and Different Machine Learning Algorithms for Detailed Tree Species Classification

“…The first is the unbalanced training samples (e.g., quality, number, and characteristics of training samples per class). The tree species with fewer samples, such as Korean pine, are often less trained or over-trained by the effects of ML algorithms, which leads to lower accuracies (RF=69.39%; SVM=57.69%) [11], [14]. The second is the patchy nature of the forest structure.…”

“…Therefore, quick and accurate tree species mapping is highly relevant for many ecological and forestry applications. methods such as Support Vector Machine (SVM) [13], Random Forest (RF) [9], [14], Extreme Gradient Boosting (XGB) [11], and K-nearest neighbor (KNN) [15]. Lim et al [16] applied RF and SVM with Hyperion and Sentinel-2 data to classify tree species and achieved overall accuracy of 0.99 and 0.97, respectively.…”

“…Wu and Zhang [17] evaluated the capability of KNN and SVM classifiers to identify combined airborne hyperspectral data of tree species with simultaneously acquired LiDAR data in southern China. The most important process in these methods is to extract features from remote sensing images, which are bands or vegetation indices highly relevant to tree species, such as SWIR bands, NDVI, and texture features, and then to input these features into the model as a mixed variable set for classification [14]. Although the mixed variable set which is beneficial to the classification of tree species is selected from a large number of bands, they all weaken the information superiority of the original data to a certain extent, which affects the potential mining of information [18].…”

Mapping Tree Species Using Advanced Remote Sensing Technologies: A State-of-the-Art Review and Perspective

“…Gillespie et al [11] proved that using Landsat OLI imagery could map species richness per ha with an accuracy of 42% (based on extracted NDVI images). However, the contribution of the moderate-resolution imagery to mapping TS is made by coupling with other sensors' data and/or other nonsensors' data, which means that the moderate-resolution imagery can aid other sensors' data to improve TS mapping accuracy, such as Landsat TM combined with HS imagery [10], Landsat OLI combined with aerial images [55], and Sentinel-2 MSI combined with VHR satellite images (GeoEye-1 and WV3) [57].…”

A Deep Fusion uNet for Mapping Forests at Tree Species Levels with Multi-Temporal High Spatial Resolution Satellite Imagery

“…In terms of methodologies, the researchers successfully employed machine learning methods such as support vector machine (SVM) and random forest (RF) to the classification of forest types based on multi-temporal satellite data and produced satisfactory classification results. Meanwhile, many comparative analyses of different machine learning methods for forest type classification based on multi-temporal data have also been conducted [14,15]. The results showed that the key factor for determining the effect of machine learning algorithm was the feature representation of the satellite data which was always in the form of manual feature extraction and optimization.…”