基于特征融合图像分割算法的生态廊道提取
Extraction of Ecological Corridor Based on Feature Fusion Image Segmentation Algorithm
葛军阳 1张宝铮1
作者信息
- 1. 长沙市规划勘测设计研究院,湖南 长沙 410007
- 折叠
摘要
为提取最短路径生态廊道且兼顾多种生物的迁徙可行性,提出基于特征融合图像分割算法的生态廊道提取方法.融合颜色特征与纹理特征得到生态廊道的感兴趣区域,构建卷积神经网络模型,将研究区景观分为林地、耕地、草地、水域等类型,据此构建路径栅格图;利用二维信息素更新策略、动态启发因子信息素因子策略改进传统蚁群算法,以栅格图为对象使用改进蚁群算法规划最短的生态廊道路径.结果表明,该方法图像分割F值在0.954～0.984 之间,波动性小;提取的生态廊道路径相对较短、拐点较少,起始点与终点之间更容易实现物质流动.
Abstract
In order to extract the shortest path ecological corridor and take into account the migration feasi-bility of various organisms,an ecological corridor extraction method based on feature fusion image segmen-tation algorithm was proposed.Color features and texture features were integrated to obtain the area of inter-est of the ecological corridor,the convolutional neural network model was built to divide the landscape of the study area into forest land,cultivated land,grassland,water and other types,and a path grid map ac-cordingly was constructed;the two-dimensional pheromone update strategy and the dynamic heuristic factor pheromone factor strategy were used to improve the traditional ant colony algorithm,the improved ant colo-ny algorithm was used to plan the shortest ecological corridor path based on a grid graph.The experimental results showed that the image segmentation F value of the modified method was between 0.954 and 0.984,and the fluctuation was small;the extracted ecological corridor path was relatively short,with fewer inflec-tion points,making it easier to realize material flow between the starting point and the end point.
关键词
特征融合/图像分割/卷积神经网络/蚁群算法/信息素/生态廊道提取Key words
feature fusion/image segmentation/convolutional neural network/ant colony algorithm/pheromone/extraction of ecological corridor引用本文复制引用
出版年
2024
NETL
NSTL
万方数据