Feature relevance analysis for 3D point cloud classification using deep learning
3D point clouds acquired by laser scanning and other techniques are difficult to interpret because of their irregular structure. To make sense of this data and to allow for the derivation of useful information, a segmentation of the points in groups, units, or classes fit for the specific use case i...
Gespeichert in:
| Hauptverfasser: | , , , |
|---|---|
| Dokumenttyp: | Kapitel/Artikel Konferenzschrift |
| Sprache: | Englisch |
| Veröffentlicht: |
29 May 2019
|
| In: |
ISPRS Geospatial Week 2019
Year: 2019, Pages: 373-380 |
| DOI: | 10.5194/isprs-annals-IV-2-W5-373-2019 |
| Online-Zugang: | Resolving-System, kostenfrei, Volltext: https://doi.org/10.5194/isprs-annals-IV-2-W5-373-2019 Verlag, kostenfrei, Volltext: https://isprs-annals.copernicus.org/articles/IV-2-W5/373/2019/4950 
|
| Verfasserangaben: | Ashutosh Kumar, Katharina Anders, Lukas Winiwarter, Bernhard Höfle |
MARC
| LEADER | 00000caa a2200000 c 4500 | ||
|---|---|---|---|
| 001 | 1891393367 | ||
| 003 | DE-627 | ||
| 005 | 20260203163248.0 | ||
| 007 | cr uuu---uuuuu | ||
| 008 | 240617s2019 xx |||||o 00| ||eng c | ||
| 024 | 7 | |a 10.5194/isprs-annals-IV-2-W5-373-2019 |2 doi | |
| 035 | |a (DE-627)1891393367 | ||
| 035 | |a (DE-599)KXP1891393367 | ||
| 035 | |a (OCoLC)1475299903 | ||
| 040 | |a DE-627 |b ger |c DE-627 |e rda | ||
| 041 | |a eng | ||
| 084 | |a 61 |2 sdnb | ||
| 100 | 1 | |a Kumar, Ashutosh |e VerfasserIn |0 (DE-588)1333279140 |0 (DE-627)1891393650 |4 aut | |
| 245 | 1 | 0 | |a Feature relevance analysis for 3D point cloud classification using deep learning |c Ashutosh Kumar, Katharina Anders, Lukas Winiwarter, Bernhard Höfle |
| 264 | 1 | |c 29 May 2019 | |
| 300 | |a 8 | ||
| 336 | |a Text |b txt |2 rdacontent | ||
| 337 | |a Computermedien |b c |2 rdamedia | ||
| 338 | |a Online-Ressource |b cr |2 rdacarrier | ||
| 500 | |a Gesehen am 17.06.2024 | ||
| 520 | |a 3D point clouds acquired by laser scanning and other techniques are difficult to interpret because of their irregular structure. To make sense of this data and to allow for the derivation of useful information, a segmentation of the points in groups, units, or classes fit for the specific use case is required. In this paper, we present a non-end-to-end deep learning classifier for 3D point clouds using multiple sets of input features and compare it with an implementation of the state-of-the-art deep learning framework PointNet++. We first start by extracting features derived from the local normal vector (normal vectors, eigenvalues, and eigenvectors) from the point cloud, and study the result of classification for different local search radii. We extract additional features related to spatial point distribution and use them together with the normal vector-based features. We find that the classification accuracy improves by up to 33% as we include normal vector features with multiple search radii and features related to spatial point distribution. Our method achieves a mean Intersection over Union (mIoU) of 94% outperforming PointNet++’s Multi Scale Grouping by up to 12%. The study presents the importance of multiple search radii for different point cloud features for classification in an urban 3D point cloud scene acquired by terrestrial laser scanning. | ||
| 700 | 1 | |a Anders, Katharina |d 1990- |e VerfasserIn |0 (DE-588)1128842580 |0 (DE-627)883601109 |0 (DE-576)48610298X |4 aut | |
| 700 | 1 | |a Winiwarter, Lukas |d 1994- |e VerfasserIn |0 (DE-588)1198882808 |0 (DE-627)1681036118 |4 aut | |
| 700 | 1 | |a Höfle, Bernhard |e VerfasserIn |0 (DE-588)1019895403 |0 (DE-627)691049297 |0 (DE-576)358986753 |4 aut | |
| 773 | 0 | 8 | |i Enthalten in |a ISPRS Geospatial Week (4. : 2019 : Enschede) |t ISPRS Geospatial Week 2019 |d [Göttingen] : [Copernicus Publications], 2019 |g (2019), Seite 373-380 |h 1 Online-Ressource |w (DE-627)1671433300 |7 nnam |
| 773 | 1 | 8 | |g year:2019 |g pages:373-380 |g extent:8 |a Feature relevance analysis for 3D point cloud classification using deep learning |
| 856 | 4 | 0 | |u https://doi.org/10.5194/isprs-annals-IV-2-W5-373-2019 |x Resolving-System |x Verlag |z kostenfrei |3 Volltext |7 0 |
| 856 | 4 | 0 | |u https://isprs-annals.copernicus.org/articles/IV-2-W5/373/2019/4950 |x Verlag |z kostenfrei |3 Volltext |7 0 |
| 951 | |a AR | ||
| 992 | |a 20240617 | ||
| 993 | |a ConferencePaper | ||
| 994 | |a 2019 | ||
| 998 | |g 1019895403 |a Höfle, Bernhard |m 1019895403:Höfle, Bernhard |d 120000 |d 120700 |e 120000PH1019895403 |e 120700PH1019895403 |k 0/120000/ |k 1/120000/120700/ |p 4 | ||
| 998 | |g 1198882808 |a Winiwarter, Lukas |m 1198882808:Winiwarter, Lukas |d 120000 |d 120700 |e 120000PW1198882808 |e 120700PW1198882808 |k 0/120000/ |k 1/120000/120700/ |p 3 | ||
| 998 | |g 1128842580 |a Anders, Katharina |m 1128842580:Anders, Katharina |d 120000 |d 120700 |e 120000PA1128842580 |e 120700PA1128842580 |k 0/120000/ |k 1/120000/120700/ |p 2 | ||
| 999 | |a KXP-PPN1891393367 |e 4539568264 | ||
| BIB | |a Y | ||
| JSO | |a {"relHost":[{"part":{"extent":"8","year":"2019","text":"(2019), Seite 373-380","pages":"373-380"},"title":[{"subtitle":"10-14 June 2019, Enschede, The Netherlands","title":"ISPRS Geospatial Week 2019","title_sort":"ISPRS Geospatial Week 2019"}],"person":[{"display":"Vosselman, George","family":"Vosselman","given":"George","role":"edt"}],"physDesc":[{"extent":"1 Online-Ressource"}],"origin":[{"publisherPlace":"[Göttingen]","dateIssuedDisp":"29 May 2019","dateIssuedKey":"2019","publisher":"[Copernicus Publications]"}],"recId":"1671433300","note":["Literaturangaben"],"name":{"displayForm":["editor(s): G. Vosselman, S.J. Oude Elberink, and M.Y. Yang"]},"disp":"ISPRS Geospatial Week (4. : 2019 : Enschede)ISPRS Geospatial Week 2019","relMultPart":[{"pubHistory":["Volume 1 (2012)-"],"note":["Gesehen am 10.11.16"],"recId":"872240061","disp":"ISPRS annals of the photogrammetry, remote sensing and spatial information sciences","part":{"number":["4, 2, W5 (2019)"],"number_sort":["4,2,W5,2019"]},"title":[{"title_sort":"ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences","title":"ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences"}],"physDesc":[{"extent":"Online-Ressource"}],"dispAlt":"ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences","origin":[{"publisherPlace":"Katlenburg-Lindau","dateIssuedDisp":"[2012]-","publisher":"Copernicus Publications"}],"id":{"eki":["872240061"],"issn":["2194-9050"],"zdb":["2874088-9"]},"corporate":[{"role":"isb","display":"International Society for Photogrammetry and Remote Sensing"}],"type":{"bibl":"serial","media":"Online-Ressource"},"language":["eng"]}],"type":{"media":"Online-Ressource","bibl":"book"},"language":["eng"],"id":{"eki":["1671433300"]},"corporate":[{"role":"aut","display":"ISPRS Geospatial Week (4., 2019, Enschede)"},{"role":"isb","display":"International Society for Photogrammetry and Remote Sensing"}]}],"title":[{"title_sort":"Feature relevance analysis for 3D point cloud classification using deep learning","title":"Feature relevance analysis for 3D point cloud classification using deep learning"}],"person":[{"family":"Kumar","display":"Kumar, Ashutosh","given":"Ashutosh","role":"aut"},{"role":"aut","family":"Anders","display":"Anders, Katharina","given":"Katharina"},{"role":"aut","family":"Winiwarter","display":"Winiwarter, Lukas","given":"Lukas"},{"family":"Höfle","display":"Höfle, Bernhard","given":"Bernhard","role":"aut"}],"physDesc":[{"extent":"8 S."}],"origin":[{"dateIssuedDisp":"29 May 2019","dateIssuedKey":"2019"}],"id":{"doi":["10.5194/isprs-annals-IV-2-W5-373-2019"],"eki":["1891393367"]},"language":["eng"],"type":{"bibl":"chapter","media":"Online-Ressource"},"recId":"1891393367","note":["Gesehen am 17.06.2024"],"name":{"displayForm":["Ashutosh Kumar, Katharina Anders, Lukas Winiwarter, Bernhard Höfle"]}} | ||
| SRT | |a KUMARASHUTFEATUREREL2920 | ||