Annotation Tool and Urban Dataset for 3D Point Cloud Semantic Segmentation

Ibrahim, Mohammed; Akhtar, Naveed; Wise, Michael J.; Mian, Ajmal

doi:10.1109/access.2021.3062547

Cited by 13 publications

(14 citation statements)

References 36 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…PointNet++ is a representative work for hierarchical networks, which uses a sampling and grouping strategy to extract the point cloud local features, the iteration of downsampling to expand the receive field of the network, and feature interpolation to finally achieve point cloud semantic segmentation. In [23], a large outdoor public dataset for 3D semantic segmentation (PC-Urban) is proposed and baseline semantic segmentation results on PC-Urban are produced by PointNet++ and PointConv. Unal et al [24] proposed a detection aware 3D semantic segmentation method which leverages localization features from an auxiliary 3D object detection task.…”

Section: Point-based Semantic Segmentation Methods For Point Cloudsmentioning

confidence: 99%

Semantic Segmentation of 3D Point Cloud Based on Spatial Eight-Quadrant Kernel Convolution

Liu

Tan

et al. 2021

Remote Sensing

View full text Add to dashboard Cite

In order to deal with the problem that some existing semantic segmentation networks for 3D point clouds generally have poor performance on small objects, a Spatial Eight-Quadrant Kernel Convolution (SEQKC) algorithm is proposed to enhance the ability of the network for extracting fine-grained features from 3D point clouds. As a result, the semantic segmentation accuracy of small objects in indoor scenes can be improved. To be specific, in the spherical space of the point cloud neighborhoods, a kernel point with attached weights is constructed in each octant, the distances between the kernel point and the points in its neighborhood are calculated, and the distance and the kernel points’ weights are used together to weight the point cloud features in the neighborhood space. In this case, the relationship between points are modeled, so that the local fine-grained features of the point clouds can be extracted by the SEQKC. Based on the SEQKC, we design a downsampling module for point clouds, and embed it into classical semantic segmentation networks (PointNet++, PointSIFT and PointConv) for semantic segmentation. Experimental results on benchmark dataset ScanNet V2 show that SEQKC-based PointNet++, PointSIFT and PointConv outperform the original networks about 1.35–2.12% in terms of MIoU, and they effectively improve the semantic segmentation performance of the networks for small objects of indoor scenes, e.g., the segmentation accuracy of small object “picture” is improved from 0.70% of PointNet++ to 10.37% of SEQKC-PointNet++.

show abstract

Section: Point-based Semantic Segmentation Methods For Point Cloudsmentioning

confidence: 99%

Semantic Segmentation of 3D Point Cloud Based on Spatial Eight-Quadrant Kernel Convolution

Liu

Tan

et al. 2021

Remote Sensing

View full text Add to dashboard Cite

show abstract

“…The second major contribution of this article is the extension of our PC-Urban dataset [21]. In this paper, we significantly extend our dataset in terms of both unlabeled raw point clouds and annotated frames.…”

Section: Proposed Datasetmentioning

confidence: 99%

“…In addition to a novel method for point-cloud semantic segmentation, this article also makes a second key contribution by extending our PC-Urban outdoor dataset [21]. The extension comes in the form of providing significantly more raw frames, as well as annotated frames.…”

Section: Introductionmentioning

confidence: 99%

“…The proposed PC-Urban_V2 dataset is collected with an Ouster LiDAR sensor installed on an SUV which drives through the downtown of Perth city, Australia. As compared to 4 billion raw points and 20 million annotated points in [21], the updated dataset (PC-Urban_V2) contains over 8 billion raw points for more than 100 K sensor frames. The dataset has now more than ∼33 K labeled instances for 25 classes comprising more than 100 million labeled data points.…”

Section: Introductionmentioning

confidence: 99%

“…The dataset has now more than ∼33 K labeled instances for 25 classes comprising more than 100 million labeled data points. We use PC-Annotate tool [21] for labeling the additional points. The updated large-scale dataset will also be made public.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Exploiting Structured CNNs for Semantic Segmentation of Unstructured Point Clouds from LiDAR Sensor

et al. 2021

Self Cite

View full text Add to dashboard Cite

Accurate semantic segmentation of 3D point clouds is a long-standing problem in remote sensing and computer vision. Due to the unstructured nature of point clouds, designing deep neural architectures for point cloud semantic segmentation is often not straightforward. In this work, we circumvent this problem by devising a technique to exploit structured neural architectures for unstructured data. In particular, we employ the popular convolutional neural network (CNN) architectures to perform semantic segmentation of LiDAR data. We propose a projection-based scheme that performs an angle-wise slicing of large 3D point clouds and transforms those slices into 2D grids. Accounting for intensity and reflectivity of the LiDAR input, the 2D grid allows us to construct a pseudo image for the point cloud slice. We enhance this image with low-level image processing techniques of normalization, histogram equalization, and decorrelation stretch to suit our ultimate object of semantic segmentation. A large number of images thus generated are used to induce an encoder-decoder CNN model that learns to compute a segmented 2D projection of the scene, which we finally back project to the 3D point cloud. In addition to a novel method, this article also makes a second major contribution of introducing the enhanced version of our large-scale public PC-Urban outdoor dataset which is captured in a civic setup with an Ouster LiDAR sensor. The updated dataset (PC-Urban_V2) provides nearly 8 billion points including over 100 million points labeled for 25 classes of interest. We provide a thorough evaluation of our technique on PC-Urban_V2 and three other public datasets.

show abstract

Human–Robot Labeling Framework to Construct Multitype Real-World Datasets

Elsharkawy

Kim

2022

IEEE Access

View full text Add to dashboard Cite

Annotation Tool and Urban Dataset for 3D Point Cloud Semantic Segmentation

Cited by 13 publications

References 36 publications

Semantic Segmentation of 3D Point Cloud Based on Spatial Eight-Quadrant Kernel Convolution

Semantic Segmentation of 3D Point Cloud Based on Spatial Eight-Quadrant Kernel Convolution

Exploiting Structured CNNs for Semantic Segmentation of Unstructured Point Clouds from LiDAR Sensor

Human–Robot Labeling Framework to Construct Multitype Real-World Datasets

Contact Info

Product

Resources

About