Vegetation plays a central role in achieving the United Nations Sustainable Development Goals (SDGs), yet its dynamics are strongly shaped by both climate change and human activities. Quantifying and separating the impacts of climate change and human activities on vegetation dynamics has remained a central challenge. However, current research still lacks precise quantitative evaluation methods to characterize the respective influence of climatic and human activities, particularly in increasingly dynamic transboundary regions. This study addresses the gap by proposing an improved threshold segmentation method through a case analysis of the Altai Mountains (AM) transboundary region. The results reveal that 21.97 % of the area showed significant restoration, while 1.05 % experienced significant degradation, with notable differences across countries. The region is undergoing warming and increased humidity, with precipitation being more strongly correlated to vegetation changes. In areas of significant restoration, human activities and climate change contributed 11.97 % and 8.74 %, respectively. The proportion of restoration driven by human activities was 14.37 % in Mongolia and 15.71 % in Russia, while it was less than 10 % in both China and Kazakhstan. Over 80 % of the restoration areas driven by climatic factors were distributed in Mongolia. In terms of the effectiveness of protected area implementation, those in China, Mongolia and Russia have all played a significant protective role. Our improved significancethreshold segmentation method proves highly effective for identifying driving factors in arid and semi-arid regions, showing great potential for broader application.