Journal of Geo-information Science >
Extraction of Urban Impervious Surface from High-Resolution Remote Sensing Imagery based on Deep Learning
Received date: 2018-12-30
Request revised date: 2019-06-12
Online published: 2019-09-24
Supported by
National Key R&D Program Projects(2016YFE0202300)
National Natural Science Foundation of China(61671332)
National Natural Science Foundation of China(41771452)
National Natural Science Foundation of China(41771454)
National Natural Science Foundation of China(51708426)
National Natural Science Foundation of China(41890820)
National Natural Science Foundation of China(41771454)
Hubei Natural Science Foundation Program Innovation Group Project(2018CFA007)
Copyright
Impervious surface is an important indicator of urban ecological environment, which is of great significance for urbanization and environmental quality assessment. The complexity of urban land use and the diversity of impervious surface materials make it a challenge to extract impervious surface directly from high-resolution remote sensing imagery. To meet the requirement of impervious surface extraction from high-resolution remote sensing imagery at the urban scale, a model of impervious surface extraction based on deep learning was proposed in this paper. Firstly, deep convolution neural network was used to extract image features. In extracting impervious surface of complex cities, a convolution layer and a pool layer were retained. While the void convolution was introduced to increase the field of receptivity and reduce the loss of information, so that each convolution output contained a larger range of information. Secondly, a probabilistic graph learning model was constructed according to its neighborhood relationship, and high-order semantic information was introduced to optimize the features to achieve accurate extraction of impervious surfaces. This paper choosed Wuhan as the experimental area, and took GaoFen-2 satellite remote sensing imagery as the data source to implement the proposed model for the extraction of impervious surface thematic information. The automatic extraction accuracy was 89.02% in the construction area and 95.55% in the urban-rural junction. Compared with the traditional machine learning algorithms such as random forest and support vector machine, the efficiency and accuracy of the proposed deep learning method were better. Statistics and analysis of the impervious surface information of the main administrative regions in Wuhan showed that the proportion of impervious surface in the whole territory of Wuhan was 11.43%, and the proportion of impervious surface in the core main urban area was close to 70%. Additionally, the present situation and development planning characteristics of Wuhan were analyzed and discussed. The impervious surface can be used as a link between urban development level and environmental quality. The distribution of impervious surface in Wuhan development planning of various administrative districts is closely related to the sustainable development of the city. Our findings suggest that the deep learning method is effective for the extraction of impervious surfaces from high-resolution remote sensing imagery, and can provide technical support and data reference for the construction of sponge city and ecological city.
CAI Bowen , WANG Shugen , WANG Lei , SHAO Zhenfeng . Extraction of Urban Impervious Surface from High-Resolution Remote Sensing Imagery based on Deep Learning[J]. Journal of Geo-information Science, 2019 , 21(9) : 1420 -1429 . DOI: 10.12082/dqxxkx.2019.180679
表1 不同方法的精度评定参数Tab. 1 Accuracy performances of the proposed method, and the RF and SVM methods |
总体精度 | Kappa系数 | 时间/s | ||||
---|---|---|---|---|---|---|
BU | RU | BU | RU | BU | RU | |
本文方法 | 0.8902 | 0.9555 | 0.8104 | 0.9167 | 58 | 33 |
随机森林 | 0.8684 | 0.9411 | 0.7761 | 0.8903 | 49 | 17 |
支持向量机 | 0.8734 | 0.9422 | 0.7846 | 0.8985 | 175 | 67 |
注:BU为建成区(Built Up area)、RU为城乡结合部(Rural Urban fringe zone)。 |
表2 武汉市7个主城区不透水面提取结果Tab. 2 Impervious extraction results of the 7 main urban areas in Wuhan (%) |
城区 | 不透水面 | 透水面 | 水面 |
---|---|---|---|
江汉区 | 65.80 | 32.72 | 1.48 |
武昌区 | 65.12 | 13.94 | 20.94 |
江岸区 | 41.66 | 35.70 | 22.65 |
青山区 | 39.15 | 44.58 | 16.27 |
汉阳区 | 34.19 | 45.27 | 20.54 |
洪山区 | 22.66 | 41.56 | 35.78 |
[1] |
|
[2] |
李德仁, 罗晖, 邵振峰 . 遥感技术在不透水层提取中的应用与展望[J]. 武汉大学学报·信息科学版, 2016,41(5):569-577.
[
|
[3] |
彭文甫, 张东辉, 何政伟 , 等. 成都市地表温度对不透水面的响应研究[J]. 遥感信息, 2010(2):98-102.
[
|
[4] |
徐涵秋 . 城市不透水面与相关城市生态要素关系的定量分析[J]. 生态学报, 2009,29(5):2456-2462.
[
|
[5] |
任鹏飞, 甘淑, 谢显奇 , 等. 基于ETM+遥感探测山地城市不透水面及其动态分析[J]. 水土保持通报, 2013,33(5):109-112.
[
|
[6] |
|
[7] |
|
[8] |
匡文慧, 刘纪远, 陆灯盛 . 京津唐城市群不透水地表增长格局以及水环境效应[J]. 地理学报, 2011,66(11):1486-1496.
[
|
[9] |
邵振峰, 潘银, 蔡燕宁 , 等. 基于Landsat年际序列影像的武汉市不透水面遥感监测[J]. 地理空间信息, 2018(1):1-5.
[
|
[10] |
|
[11] |
|
[12] |
|
[13] |
|
[14] |
|
[15] |
|
[16] |
|
[17] |
|
[18] |
刘建伟, 刘媛, 罗雄麟 . 深度学习研究进展[J]. 计算机应用研究, 2014,31(7):1921-1930.
[
|
[19] |
|
[20] |
王相海, 明莹, 方玲玲 . 概率图模型及其图像与视频应用研究[J]. 中国图象图形学报, 2009,14(9):1712-1720.
[
|
[21] |
刘波, 张源, 程涛 , 等. 基于高分二号卫星影像的城市不透水面提取[J]. 地理信息世界, 2017,24(2):103-107.
[
|
[22] |
|
[23] |
陈斌, 王宏志, 徐新良 , 等. 武汉市主城区用地结构遥感监测及动态变化[J]. 地球信息科学学报, 2019,21(2):293-303.
[
|
[24] |
姜芊孜, 王志芳, 李明翰 . 不透水地表对城市水环境的负面影响及缓解措施[J]. 城市问题, 2014(11):50-55.
[
|
[25] |
董欣, 杜鹏飞, 李志一 , 等. SWMM模型在城市不透水区地表径流模拟中的参数识别与验证[J]. 环境科学, 2008,29(6):1495-1501.
[
|
[26] |
班玉龙, 孔繁花, 尹海伟 , 等. 土地利用格局对SWMM模型汇流模式选择及相应产流特征的影响[J]. 生态学报, 2016,36(14):4317-4326.
[
|
[27] |
|
[28] |
袁修柳, 杨昆, 王保云 . 武汉市2015年不透水面的提取及其空间自相关分析[J]. 湖北农业科学, 2017,56(9):1642-1647.
[
|
[29] |
罗智勇, 王明珠, 陈婉佳 . 城市不透水面信息提取及水环境效应分析——以武汉市江南三区为例[J]. 信阳师范学院学报(自然科学版), 2016(1):148-151.
[
|
/
〈 | 〉 |