Research on Urban Tree Classification Method Based on YOLO-CNGD

cunjin zhang¹mei liu¹xinglong liu²zhixin gu^1*

• ¹Northeast Forestry University, Harbin, China
• ²Northeast Forestry University College of Forestry, Harbin, China

Urban tree Accurate classification is an important indicator for monitoring urban landscape resources and is crucial for the healthy development of urban ecosystems. Using Google Earth Engine remote sensing images, a remote sensing image dataset of urban trees in Harbin is constructed. Using YOLOv11n as the base model.tree species is fundamental for urban green space management and ecological assessment. To enhance the model's ability to capture key features, the CBAM attention mechanism is introduced in conjunction with dual channels to improve accuracy. To improve the detection accuracyaddress the challenges of small objects, the NWD loss function and overlapping tree crown detection in high-resolution remote sensing imagery, this study proposes YOLO-CNGD, a novel framework based on YOLOv11n. The key enhancements include the integration of the Convolutional Block Attention Module (CBAM) for refined feature representation, the adoption of the Normalized Wasserstein Distance (NWD) loss for robust small-object localization, the incorporation of Deformable Convolution v3 module are introduced. The ordinary convolution in the C3k2 module is replaced(DCNv3) to adapt to irregular shapes, and the replacement of standard convolutions with GhostConv to reduce the model parameters.for a lightweight design. Experiments have proven that the improved YOLO-CNGD model effectively enhances the accuracy of urban tree classification. Experiments have proven that the improved YOLO-CNGD model effectively enhances the accuracy of on a self-built urban tree classification. The dataset show that YOLO-CNGD achieves a precision rate reachedof 94.8%, thea recall rate reached 89of 91.1%, and thean mAP@0.5 reachedof 93.7%.The model balances accuracy and efficiency, showing great potential for large-scale automated urban tree inventory.