Enhanced DeepLabV3+ with OBIA and Lightweight Attention for Accurate and Efficient Tree Species Classification in UAV Images
Xue Cheng, Jianjun Chen, Junji Li, Jiayuan Yin, Qingmin Cheng, Zizhen Chen, Xinhong Li, Haotian You, Xiaowen Han, Guoqing Zhou, Adam Świtoński, Konrad W Wojciechowski, Henryk Josiński · 2025
Read Enhanced DeepLabV3+ with OBIA and Lightweight Attention for Accurate and Efficient Tree Species Classification in UAV Images by Xue Cheng, Jianjun Chen, Junji Li, Jiayuan Yin, Qingmin Cheng, Zizhen Chen, Xinhong Li, Haotian You, Xiaowen Han, Guoqing Zhou, Adam Świtoński, Konrad W Wojciechowski, Henryk Josiński (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/pmc-12736519.txt.
Description
Xue Cheng and a large team of co-authors improved drone-based tree species mapping, in a 2025 paper. Fine drone photos can sort species, but look-alike canopies confuse models. The team enhanced a standard segmentation network. Object-based grouping first tidied canopy boundaries. Feature selection kept the most telling signals, from spectra to position.
Training labels came from a forest classifier checked by eye. A light attention module then focused the network on key regions. Parameter count fell by over three quarters. Accuracy reached about ninety-five percent on their tests. The score beat the base model and two rival networks. Gains held while the model shrank, which aids field use. The authors offer the method for carbon and biodiversity monitoring. Forest managers need exactly such light, accurate tools.
Catalog record
| CREATOR | Xue Cheng, Jianjun Chen, Junji Li, Jiayuan Yin, Qingmin Cheng, Zizhen Chen, Xinhong Li, Haotian You, Xiaowen Han, Guoqing Zhou, Adam Świtoński, Konrad W Wojciechowski, Henryk Josiński |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 53 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |