Multispectral airborne laser scanning for tree species classification: A benchmark of machine learning and deep learning algorithms

Climate-smart and biodiversity-preserving forestry demands precise information on forest resources, extending to the individual tree level. Multispectral airborne laser scanning (ALS) has shown promise in automated point cloud processing, but challenges remain in leveraging deep learning techniques and identifying rare tree species in class-imbalanced datasets. This study addresses these gaps by conducting a comprehensive benchmark of deep learning and traditional shallow machine learning methods for tree species classification. For the study, we collected high-density multispectral ALS data (>1000 pts/m(2)) at three wavelengths using the FGI-developed HeliALS system, complemented by existing Optech Titan data (35 pts/m(2)), to evaluate the species classification accuracy of various algorithms in a peri-urban study area located in southern Finland. We established a field reference dataset of 6326 segments across nine species using a newly developed browser-based crowdsourcing tool, which facilitated efficient data annotation. The ALS data, including a training dataset of 1065 segments, was shared with the scientific community to foster collaborative research and diverse algorithmic contributions. Based on 5261 test segments, our findings demonstrate that point-based deep learning methods, particularly a point transformer model, outperformed traditional machine learning and image-based deep learning approaches on high-density multispectral point clouds. For the high-density ALS dataset, a point transformer model provided the best performance reaching an overall (macro-average) accuracy of 87.9% (74.5%) with a training set of 1065 segments and 92.0% (85.1%) with a larger training set of 5000 segments. With 1065 training segments, the best image-based deep learning method, DetailView, reached an overall (macro-average) accuracy of 84.3% (63.9%), whereas a shallow random forest (RF) classifier achieved an overall (macro-average) accuracy of 83.2% (61.3%). For the sparser ALS dataset, an RF model topped the list with an overall (macro-average) accuracy of 79.9% (57.6%), closely followed by the point transformer at 79.6% (56.0%). Importantly, the overall classification accuracy of the point transformer model on the HeliALS data increased from 73.0% with no spectral information to 84.7% with single-channel reflectance, and to 87.9% with spectral information of all the three channels. Furthermore, we studied the scaling of the classification accuracy as a function of point density and training set size using 5-fold cross-validation of our dataset. Based on our findings, multispectral information is especially beneficial for sparse point clouds with 1-50 pts/m(2). Furthermore, we observed that the classification error follows a power law "(m) approximate to m(-alpha )as a function of the training set size m, and the classification error of the point transformer reduced significantly faster with increasing training set size compared to RF.