Tree species’ composition of forests is essential in forest management and nature conservation. We aimed to identify the tree species structure of a floodplain forest area using a hyperspectral image. We proposed an efficient novel strategy including the testing of three dimension reduction (DR) methods: Principal Component Analysis, Minimum Noise Fraction (MNF) and Indipendent Component Analysis with five machine learning (ML) algorithms (Maximum Likelihood Classifier, Support Vector Classification, Support Vector Machine, Random Forest and Artificial Neural Network) to find the most accurate outcome; altogether 300 models were calculated. Post-classification was applied by combining the multiresolution segmentation and filtering. MNF was the most efficient DR technique, and at least 7 components were needed to gain an overall accuracy (OA) of > 75%. Forty-five models had > 80% OAs; MNF was 43, and the Maximum Likelihood was 19 times among these models. Best classification belonged to MNF with 10 components and Maximum Likelihood classifier with the OA of 83.3%. Post-classification increased the OA to 86.1%. We quantified the differences among the possible DR and ML methods, and found that even > 10% worse model can be found using popular standard procedures related to the best results. Our workflow calls the attention of careful model selection to gain accurate maps.

Species composition of forests is a very important component from the point of view of nature conservation and forestry. We aimed to identify 10 tree species in a hilly forest stand using a hyperspectral aerial image with a particular focus on two invasive species, namely Ailanthus tree and black locust. Deep learning‐based training data augmentation (TDA) and post‐classification techniques were tested with Random Forest and Support Vector Machine (SVM) classifiers. SVM had better performance with 81.6% overall accuracy (OA). TDA increased the OA to 82.5% and post‐classification with segmentation improved the total accuracy to 86.2%. The class‐level performance was more convincing: the invasive Ailanthus trees were identified with 40% higher producer's and user's accuracies (PA and UA) to 70% related to the common technique (using a training dataset and classifying the trees). The PA and UA did not change in the case of the other invasive species, black locust. Accordingly, this new method identifies well Ailanthus, a sparsely distributed species in the area; while it was less efficient with black locust that dominates larger patches in the stand. The combination of the two ancillary steps of hyperspectral image classification proved to be reasonable and can support forest management planning and nature conservation in the future.

No abstract