摘要
为了提高城市高分辨率遥感影像中植被信息的提取精度,提出一种改进的全卷积神经网络模型,通过大量的训练数据获得最佳模型参数,进行植被信息的提取,并与支持向量机、面向对象法、经典的FCN模型方法提取的植被信息进行对比分析。结果表明:提出的网络模型不但能够有效缓解"椒盐现象",还能保证小面积的植被提取与植被区域边界的准确性。该方法可自动综合多种特征,所以可有效减少植被像元的误分与漏分现象,提高植被提取精度。
In order to improve the extraction of urban vegetation from high-resolution remote sensing image, an novel full convolutional neural network model was proposed. The best model parameters were obtained through a large amount of training data, and the vegetation information was extracted. The vegetation information extracted by support vector machine, object-oriented method and classical FCN model method was compared and analyzed. The results show that the proposed network model can not only effectively alleviate the "salt and pepper phenomenon", but also ensure the accuracy of small-area vegetation extraction and vegetation area boundaries. The method can automatically integrate multiple features, so it can effectively reduce the misclassification and leakage of vegetation pixels and improve the accuracy of vegetation extraction.
作者
马海艺
张天怡
代沁伶
代飞
王雷光
Ma Haiyi;Zhang Tianyi;Dai Qinling;Dai Fei;Wang Leiguang(College of Forestry, Southwest Forestry University, Kunming Yunnan 650233, China;Key Laboratory of National Forestry and Grassland Administration for Forestry and Ecological Big Data, Southwest Forestry University, Kunming Yunnan 650233, China;Institute of Big Data and Artificial Intelligence, Southwest Forestry University, Kunming Yunnan 650233, China;College of Design, Southwest Forestry University, Kunming Yunnan 650233, China)
出处
《西南林业大学学报（自然科学）》
CAS
北大核心
2019年第3期117-123,共7页
Journal of Southwest Forestry University：Natural Sciences
基金
国家自然科学基金项目(41571372)资助
关键词
全卷积神经网络
高分辨率
遥感影像
城市植被
椒盐现象
fully convolutional networks
high-resolution
remote sensing image
urban vegetation
salt-andpepper phenomenon
