Related Experiment Video
Updated: Sep 18, 2025
03:29
Author Spotlight: Impact of Physical Barriers on Rodent Populations in Farmland Areas
Published on: March 8, 2024
697
Method for assessing rodent infestation in plateau based on SegFormer.
Xiangjie Huang1,2, Guoying Zhang3, Chunmei Li1,2,4
1School of Computer Technology and Application, Qinghai University, Xining, Qinghai, China.
Plos One
|June 26, 2025
Summary
Detecting rodent burrows on plateaus is crucial for grassland health. This study introduces EM-SegFormer, an enhanced AI model using drone imagery for accurate rodent infestation mapping, aiding grassland restoration efforts.
Area of Science:
- Ecology
- Remote Sensing
- Artificial Intelligence
Background:
- Rodent infestation severely degrades grassland ecosystems, particularly on plateaus.
- Effective detection of rodent burrows is essential for ecological assessment and management.
- Current methods lack the precision needed for complex highland environments.
Purpose of the Study:
- To develop and validate an advanced semantic segmentation model for detecting rodent burrows in highland areas.
- To improve the accuracy of rodent infestation mapping using aerial imagery.
- To provide a robust technical solution for grassland restoration and management.
Main Methods:
- Utilized unmanned aerial vehicle (UAV) for high-resolution video data acquisition.
- Constructed a labeled dataset of rodent burrows from processed aerial imagery.
- Modified the SegFormer model by integrating an efficient multi-scale attention (EMA) mechanism and a multi-kernel convolution feed-forward network (MCFN), creating EM-SegFormer.
Main Results:
- The proposed EM-SegFormer model demonstrated effective performance in segmenting rodent burrows.
- The integrated EMA and MCFN mechanisms enhanced the model's ability to handle small targets and complex backgrounds.
- The method achieved good results on the specialized rodent burrows dataset.
Conclusions:
- EM-SegFormer offers a novel and effective approach for detecting rodent infestation in plateau environments.
- This AI-driven method provides reliable technical support for ecological monitoring and grassland restoration.
- Accurate rodent burrow detection is key to mitigating grassland degradation and preserving ecosystem health.
