product.page.titleprefix
PalLoc6D - Estimating the Pose of a Euro Pallet with an RGB Camera based on Synthetic Training Data

Simple item page
cris.virtual.author-gnd1316353338
cris.virtual.author-gnd1199320811
cris.virtual.author-gnd#PLACEHOLDER_PARENT_METADATA_VALUE#
cris.virtual.author-gnd1199294179
cris.virtual.author-gnd1199292346
cris.virtual.author-orcid0000-0002-4051-2268
cris.virtual.author-orcid0000-0002-1687-7654
cris.virtual.author-orcid0000-0001-8151-3067
cris.virtual.author-orcid0000-0001-9823-7679
cris.virtual.author-orcid0000-0003-3648-576X
cris.virtualsource.author-gnd00a2fa70-2fb7-48d9-88e3-4b11fa81b672
cris.virtualsource.author-gnd71f56652-a55c-487b-8191-00e9fe67be11
cris.virtualsource.author-gnd36a4fab6-5103-42c0-8f34-1dce86ea3f72
cris.virtualsource.author-gndf9f01a31-5d87-4e54-b954-f35b36352d46
cris.virtualsource.author-gndd6fbdc42-0742-4a5f-93a0-3486f85934fe
cris.virtualsource.author-orcid00a2fa70-2fb7-48d9-88e3-4b11fa81b672
cris.virtualsource.author-orcid71f56652-a55c-487b-8191-00e9fe67be11
cris.virtualsource.author-orcid36a4fab6-5103-42c0-8f34-1dce86ea3f72
cris.virtualsource.author-orcidf9f01a31-5d87-4e54-b954-f35b36352d46
cris.virtualsource.author-orcidd6fbdc42-0742-4a5f-93a0-3486f85934fe
datacite.relation.Referenceshttps://github.com/NVIDIA/Dataset_Synthesizer
datacite.relation.Referenceshttps://github.com/NVlabs/Deep_Object_Pose
datacite.resourceTypeResearchData
datacite.resourceTypeGeneralDataset
dc.contributor.authorKnitt, Markus
dc.contributor.authorSchyga, Jakob
dc.contributor.authorAdamanov, Asan
dc.contributor.authorHinckeldeyn, Johannes
dc.contributor.authorKreutzfeldt, Jochen
dc.date.accessioned2022-08-26T10:26:21Z
dc.date.available2022-08-26T10:26:21Z
dc.date.issued2022-08-26
dc.description.abstractPalLoc6D contains 50 000 synthetically generated images of a photorealistic pallet in a domain-radomized environment. PalLoc6D includes annotations of the pallets' 6D pose. The data was created using the NVIDIA Dataset Synthesizer (NDDS, https://github.com/NVIDIA/Dataset_Synthesizer). Additionally, a photorealistic 3D model of a Euro pallet is provided. PalLoc6D can be used to train neural networks for RGB camera-based 6D pallet pose estimation, such as Nvidia's "Deep Object Pose Estimation" (DOPE, https://github.com/NVlabs/Deep_Object_Pose). Furthermore, the weights of the DOPE algorithm, trained with the annotated images are included in PalLoc6D to allow a quick start for experimenting with 6D pose estimation. PalLoc6D was published as part of the paper "Estimating the Pose of a Euro Pallet with an RGB Camera based on Synthetic Training Data", which was presented at the WGTL Fachkolloquium 2022 in Bremen and will be subsequently published in the Logistics Journal. The purpose, creation, and validation of the dataset are further elaborated in the publication. Paper abstract: "Estimating the pose of a pallet and other logistics objects is crucial for various use cases, such as automatized material handling or tracking. Innovations in computer vision, computing power, and machine learning open up new opportunities for device-free localization based on cameras and neural networks. Large image datasets with annotated poses are required for training the network. Manual annotation, especially of 6D poses, is an extremely labor-intensive process. Hence, newer approaches often leverage synthetic training data to automatize the process of generating annotated image datasets. In this work, the generation of synthetic training data for 6D pose estimation of pallets is presented. The data is then used to train the Deep Object Pose Estimation (DOPE) algorithm. The experimental validation of the algorithm proves that the 6D pose estimation of a standardized Euro pallet with a Red-Green-Blue (RGB) camera is feasible. The comparison of the results from three varying datasets under different lighting conditions shows the relevance of an appropriate dataset design to achieve an accurate and robust localization. The quantitative evaluation shows an average position error of less than 20 cm for the preferred dataset. The validated training dataset and a photorealistic model of a Euro pallet are publicly provided."en
dc.identifier.doi10.15480/336.4470
dc.identifier.urihttp://hdl.handle.net/11420/13164
dc.language.isoende_DE
dc.rights.urihttps://creativecommons.org/publicdomain/zero/1.0/de_DE
dc.subject6D pose estimationde_DE
dc.subjectEuro palletde_DE
dc.subjectsynthetic training datasetde_DE
dc.subjectRGB camerade_DE
dc.subjectDeep Object Pose Estimationde_DE
dc.subjectNVIDIA Dataset Synthesizerde_DE
dc.subject.ddc004: Informatikde_DE
dc.subject.ddc600: Technikde_DE
dc.subject.ddc620: Ingenieurwissenschaftende_DE
dc.titlePalLoc6D - Estimating the Pose of a Euro Pallet with an RGB Camera based on Synthetic Training Datade_DE
dc.typeDatasetde_DE
dc.type.casraiOther
dc.type.diniResearchData
dc.type.driverother
dcterms.DCMITypeDataset
dspace.entity.typeProduct
local.researchdata.deleteaftertenyearstruede_DE
local.type.legacyDataset
tuhh.abstract.englishPalLoc6D contains 50 000 synthetically generated images of a photorealistic pallet in a domain-radomized environment. PalLoc6D includes annotations of the pallets' 6D pose. The data was created using the NVIDIA Dataset Synthesizer (NDDS, https://github.com/NVIDIA/Dataset_Synthesizer). Additionally, a photorealistic 3D model of a Euro pallet is provided. PalLoc6D can be used to train neural networks for RGB camera-based 6D pallet pose estimation, such as Nvidia's "Deep Object Pose Estimation" (DOPE, https://github.com/NVlabs/Deep_Object_Pose). Furthermore, the weights of the DOPE algorithm, trained with the annotated images are included in PalLoc6D to allow a quick start for experimenting with 6D pose estimation. PalLoc6D was published as part of the paper "Estimating the Pose of a Euro Pallet with an RGB Camera based on Synthetic Training Data", which was presented at the WGTL Fachkolloquium 2022 in Bremen and will be subsequently published in the Logistics Journal. The purpose, creation, and validation of the dataset are further elaborated in the publication. Paper abstract: "Estimating the pose of a pallet and other logistics objects is crucial for various use cases, such as automatized material handling or tracking. Innovations in computer vision, computing power, and machine learning open up new opportunities for device-free localization based on cameras and neural networks. Large image datasets with annotated poses are required for training the network. Manual annotation, especially of 6D poses, is an extremely labor-intensive process. Hence, newer approaches often leverage synthetic training data to automatize the process of generating annotated image datasets. In this work, the generation of synthetic training data for 6D pose estimation of pallets is presented. The data is then used to train the Deep Object Pose Estimation (DOPE) algorithm. The experimental validation of the algorithm proves that the 6D pose estimation of a standardized Euro pallet with a Red-Green-Blue (RGB) camera is feasible. The comparison of the results from three varying datasets under different lighting conditions shows the relevance of an appropriate dataset design to achieve an accurate and robust localization. The quantitative evaluation shows an average position error of less than 20 cm for the preferred dataset. The validated training dataset and a photorealistic model of a Euro pallet are publicly provided."de_DE
tuhh.gvk.hasppnfalse
tuhh.hasurnfalse
tuhh.identifier.doi10.15480/336.4470
tuhh.identifier.urnurn:nbn:de:gbv:830-882.0191325
tuhh.publication.instituteTechnische Logistik W-6de_DE
tuhh.type.opusDataset
tuhh.type.rdmtrue

Files

Original bundle
Now showing1 - 5 of 6
No Thumbnail Available
Name:
PalLoc6D_part2.zip
Size:
4.7 GB
Format:
ZIP Archive
Description:
Training Data Part 2
Download
No Thumbnail Available
Name:
PalLoc6D_part4.zip
Size:
4.72 GB
Format:
ZIP Archive
Description:
Training Data Part 4
Download
No Thumbnail Available
Name:
ndds3_pallet.pth
Size:
191.82 MB
Format:
PTH is a data file for Machine Learning with PyTorch. PyTorch is an open source machine learning library based on the Torch library. It is primarily developed by Facebooks artificial intelligence research group.
Download
No Thumbnail Available
Name:
PalLoc6D_part1.zip
Size:
4.74 GB
Format:
ZIP Archive
Description:
Training Data Part 1
Download
No Thumbnail Available
Name:
Euro Pallet 3D Model.zip
Size:
3.5 GB
Format:
ZIP Archive
Description:
Photorealistic model of a Euro pallet
Download
License bundle
Now showing1 - 1 of 1
No Thumbnail Available
Name:
license.txt
Size:
3.12 KB
Format:
Item-specific license agreed upon to submission
Description:
Download

Collections

Research Data