Creating a land use/land cover dictionary based on multiple pairs of OSM and reference datasets
This is a modal window.
The media could not be loaded, either because the server or network failed or because the format is not supported.
Formal Metadata
Title |
| |
Title of Series | ||
Number of Parts | 351 | |
Author | ||
Contributors | ||
License | CC Attribution 3.0 Unported: You are free to use, adapt and copy, distribute and transmit the work or content in adapted or unchanged form for any legal purpose as long as the work is attributed to the author in the manner specified by the author or licensor. | |
Identifiers | 10.5446/68905 (DOI) | |
Publisher | ||
Release Date | ||
Language | ||
Production Year | 2022 |
Content Metadata
Subject Area | ||
Genre | ||
Abstract |
| |
Keywords |
00:00
Data dictionaryInformation engineeringObservational studyMathematical analysisMultiplicationGoodness of fitReference dataCovering spaceComputer animation
00:13
PlanningTransportation theory (mathematics)InformationTexture mappingResource allocationCovering spaceService (economics)EstimationTelecommunicationBit rateData dictionaryObservational studyCorrespondence (mathematics)ForestIntegrated development environmentAngular resolutionTemporal logicShape (magazine)PolygonInflection pointBest, worst and average caseMaxima and minimaSample (statistics)Total S.A.CalculationBuildingArtificial neural networkMathematical analysisCountingRankingComplete metric spaceMetropolitan area networkConsistencyObject (grammar)Database transactionPerformance appraisalPairwise comparisonObservational studyMultiplicationMaxima and minimaObject (grammar)Sample (statistics)Best, worst and average caseData dictionaryMappingCorrespondence (mathematics)InformationMereologyCartesian coordinate systemInternet service providerTexture mappingConsistencyWave packetForestResultantMeasurementCalculationThresholding (image processing)Different (Kate Ryan album)Limit (category theory)SpacetimeReference dataSystem of linear equationsCountingSoftware testingCovering spaceMathematical analysisSet (mathematics)Latent heatMetropolitan area networkCASE <Informatik>Open setComputer animationTable
Transcript: English(auto-generated)
00:00
Good afternoon, I'm Yao Ming and I will present your work on creating a land-use land-cover dictionary based on multiple pairs of OSM and reference datasets. Land-use land-cover map has many significant applications. OpenStreetMap can provide useful information for land-use land-cover mapping.
00:24
The classifications for OSM and land-use land-cover are different, so a correspondence between them is needed. In recent studies, correspondences are mostly established in two methods, subjectively or automatically based on one pair of OSM and reference data.
00:41
Both methods may lead to a problem, that is, the correspondence based on one study area may not always be applicable to others. So, we proposed an approach to use multiple pairs of OSM and reference data for creating a dictionary. Our study area includes 50 pan-European metropolitans, and the reference data is urban atlas.
01:05
For each study area, all the OSM tags are intersected with different land-use land-cover classes, and the class with the maximum intersecting area is viewed as the most appropriate class for this OSM tag. Besides, four measures are designed to describe the dictionary.
01:21
The first one is sample count, which denotes how frequent an OSM tag is appeared in different study areas. The second one is average area percentage, which denotes the average of the area percentages of an OSM tag in multiple OSM datasets. The third one is sample ratio, which denotes the percentage of study areas that an OSM tag corresponds to a specific reference class.
01:48
The fourth one is average maximum percentage. Maximum percentage denotes the ratio of the maximum intersecting area for a pair of OSM tag and reference class to the total area for this OSM tag.
02:03
We calculate the four measures for each pair of OSM and land-use land-cover class. Then, all the calculation results are counted to establish the dictionary. This is the dictionary we established. Only a part is shown due to space limitations. In the analysis, we found that more than 1000 OSM tags have a low sample count, and only 37 OSM tags had a relatively large average area percentage.
02:32
These two findings mean that only a small proportion of OSM tags can play a decisive role in the dictionary. And how can we use the dictionary? There are three possible applications.
02:43
The first one is increasing land-use land-cover mapping accuracy. For example, in scenario 1, we use all the OSM tags in dictionary for mapping. And in scenario 2, we set screening threshold for three measures, and then accuracy was increased for all the test areas.
03:02
The second application is picking up training data. For example, OSM tag forest has high sample ratio and average maximum percentage with land-use land-cover class forests, which indicates a high consistency. In this case, the OSM objects tagged forest could be used for picking up forest training data.
03:23
The third application is OSM quality assessment. For example, a low average maximum percentage indicates a low consistency between OSM tag and reference class. And we can detect potential incorrect tags in this OSM tag. In conclusion, this study proposed an approach to establish a dictionary based on multiple pairs of OSM and reference datasets.
03:48
And it is beneficial to create such a dictionary for essential applications. Thanks for your attention.