Fully Convolutional Neural Network for Impervious Surface Segmentation in Mixed Urban Environment

The urgency of creating appropriate, high-resolution data products such as impervious cover information has increased as cities face rapid growth as well as climate change and other environmental challenges. This work explores the use of fully convolutional neural networks (FCNNs )—specifically UNet with a ResNet-152 encoder—in mapping impervious surfaces at the pixel level from WorldView-2 in a mixed urban/residential environment. We investigate three-, four-, and eight-band multispectral inputs to the FCNN. Resulting maps are promising in both qualitative and quantitative assessment when compared to automated land use/land cover products. Accuracy was assessed by F1 and average precision (AP) scores, as well as receiver operating characteristic curves, with area under the curve (AUC ) used as an additional accuracy metric. The four-band model shows the highest average test-set accuracies (F1, AP, and AUC of 0.709, 0.82, and 0.807, respectively), with higher AP and AUC than the automated land use/land cover products, indicating the utility of the blue-green-red-infrared channels for the FCNN. Improved performance was seen in residential areas, with worse performance in more densely developed areas.

Download Full-text

Land use changes in the hinterland of major Slovak cities

Geografie ◽

10.37040/geografie2010115040413 ◽

2010 ◽

Vol 115 (4) ◽

pp. 413-439 ◽

Cited By ~ 4

Author(s):

Martin Šveda ◽

Daniela Vigašová

Keyword(s):

Land Use ◽

Land Cover ◽

Spatial Structure ◽

Land Use Changes ◽

Detailed Comparison ◽

Land Use Land Cover ◽

Residential Areas ◽

Shopping Centres ◽

Agricultural Use ◽

Land Types

The countryside around major Slovak cities is undergoing significant transformation. The construction of shopping centres, administrative buildings, logistical sites, residential areas and changes in the agricultural use of land are causing vast changes in land use (land cover). The objective of this paper is to examine changes in the spatial structure of land use in the hinterland of 11 Slovak cities, with more than 50 thousand inhabitants, during the period from 2000 to 2008. On the basis of a detailed comparison of data obtained from the Aggregated Areas of Land Types database (Úhrnné hodnoty druhov pozemkov) we analyzed changes in land use in 847 municipalities within the Functional Urban Regions of Bratislava, Košice, Prešov, Nitra, Žilina, Banská Bystrica, Trnava, Trenčín, Poprad and Prievidza. The results of the research confirmed significantly differentiated changes in land use. Whereas in the capital of Bratislava changes in land use are primarily caused by suburbanization, creating a relatively compact suburban zone, changes in land use were recorded only in selected sites in the rest of these major Slovak cities.

Download Full-text

Quantifying Land Use/Land Cover and Landscape Pattern Changes and Impacts on Ecosystem Services

International Journal of Environmental Research and Public Health ◽

10.3390/ijerph17010126 ◽

2019 ◽

Vol 17 (1) ◽

pp. 126 ◽

Cited By ~ 5

Author(s):

Qingjian Zhao ◽

Zuomin Wen ◽

Shulin Chen ◽

Sheng Ding ◽

Minxin Zhang

Keyword(s):

Land Use ◽

Land Cover ◽

Landscape Heterogeneity ◽

Landscape Patterns ◽

Water Bodies ◽

Guizhou Province ◽

Land Use Land Cover ◽

Residential Areas ◽

Regulatory Policies ◽

Land Types

Based on satellite remote sensing image, GIS and Fragstats, this study modeled and calculated the dynamic changes of land use, land cover and landscape patterns in Guizhou Province, China, and calculated the changes of ecosystem service values (ESVs). The impacts of the evolution of landscape patterns on the ESVs were analyzed, and reasonable policy recommendations were made. The findings are as follows: (1) In the past two decades, the area of cropland and grassland has decreased; the area of water bodies, urban and rural, industrial and mining, and residential areas has increased; the area of forestland has increased first and then decreased. (2) The two major types of landscapes, cropland and grassland, are clearly being replaced by two land types, forest land and water bodies. (3) Overall, the degree of landscape aggregation and adjacency has decreased, and the landscape heterogeneity has increased. (4) The total amount of ESV in 2000, 2008, 2013 and 2017 was 2574 × 108 Yuan RMB, 2605 × 108 Yuan RMB, 2618 × 108 Yuan RMB and 2612 × 108 Yuan RMB, respectively. The changes of landscape patterns had important impacts on the ESVs. In order to solve the problems caused by the increasingly prominent changes in the landscape patterns and improve the ESVs, it is necessary to rationally plan and allocate land resources, optimize the industrial structures, and develop effective regulatory policies.

Download Full-text

Land use/land cover mapping using multi-scale texture processing of high resolution data

IOP Conference Series Earth and Environmental Science ◽

10.1088/1755-1315/18/1/012185 ◽

2014 ◽

Vol 18 ◽

pp. 012185

Author(s):

S N Wong ◽

M L R Sarker

Keyword(s):

Land Use ◽

High Resolution ◽

Land Cover ◽

Land Cover Mapping ◽

Land Use Land Cover ◽

High Resolution Data ◽

Multi Scale ◽

Resolution Data ◽

Texture Processing

Download Full-text

Landscape Influences on Stream Habitats and Biological Assemblages

Landscape Influences on Stream Habitats and Biological Assemblages ◽

10.47886/9781888569766.ch28 ◽

2006 ◽

Keyword(s):

Land Use ◽

Land Cover ◽

Benthic Invertebrate ◽

Base Flow ◽

Flow Index ◽

Physical Habitat ◽

Land Use Land Cover ◽

Surficial Geology ◽

Impervious Cover ◽

Base Flow Index

<em>Abstract.</em>—–We demonstrate the effects of percent impervious cover (PIC) on biophysical properties of Lake Ontario tributary streams. Biophysical data (fish assemblages, benthic invertebrate assemblages (benthos), instream physical habitat, and temperature) were collected from more than 575 wadeable stream sites. A geographic information system application was developed to characterize the landscape upstream of each site (i.e., drainage area, surficial geology, land use/land cover, slope, stream length, and climate). Total PIC of catchments was estimated from land use/land cover, and a base flow index was derived from the surficial geology. The relationship between PIC and biophysical responses was determined after statistically removing the effects of natural landscape features (i.e., catchment area, slope, base flow index) on those responses. Contrasts in PIC from natural conditions (<3% to 10%) were related to variations in fish and benthos assemblages. Both coldwater sensitive and warmwater tolerant fish and diverse benthos assemblages were found in catchments with low PIC. At more than 10 PIC (i.e., about 50% urban), both fish and benthos consisted of mainly warmwater or tolerant assemblages. For example, trout were absent and minnows were dominant. While some of the apparent PIC effect may have been confounded by land use/land cover and surficial geology, the consistency of the findings even after natural catchment conditions were considered suggests that the threshold response is valid. Percent impervious cover had a weaker effect on instream geomorphic variables than on biological variables. The models derived from this study can be used to predict stream biophysical conditions for catchments with varying levels of development.

Download Full-text

Hyperspectral Data for Land use/Land cover classification

ISPRS - International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences ◽

10.5194/isprsarchives-xl-8-991-2014 ◽

2014 ◽

Vol XL-8 ◽

pp. 991-995 ◽

Cited By ~ 1

Author(s):

D. Vijayan ◽

G. Ravi Shankar ◽

T. Ravi Shankar

Keyword(s):

Land Use ◽

High Resolution ◽

Land Cover ◽

Classification Systems ◽

Land Use Land Cover ◽

High Resolution Data ◽

Liss Iii ◽

Hyperion Image ◽

Resolution Data

An attempt has been made to compare the multispectral Resourcesat-2 LISS III and Hyperion image for the selected area at sub class level classes of major land use/ land cover. On-screen interpretation of LISS III (resolution 23.5 m) was compared with Spectral Angle Mapping (SAM) classification of Hyperion (resolution 30m). Results of the preliminary interpretation of both images showed that features like fallow, built up and wasteland classes in Hyperion image are clearer than LISS-III and Hyperion is comparable with any high resolution data. Even canopy types of vegetation classes, aquatic vegetation and aquatic systems are distinct in Hyperion data. Accuracy assessment of SAM classification of Hyperion compared with the common classification systems followed for LISS III there was no much significant difference between the two. However, more number of vegetation classes could be classified in SAM. There is a misinterpretation of built up and fallow classes in SAM. The advantages of Hyperion over visual interpretation are the differentiation of the type of crop canopy and also crop stage could be confirmed with the spectral signature. The Red edge phenomenon was found for different canopy type of the study area and it clearly differentiated the stage of vegetation, which was verified with high resolution image. Hyperion image for a specific area is on par with high resolution data along with LISS III data.

Download Full-text