Using Deep Learning in Semantic Classification for Point Cloud Data

3D point cloud simplification is an important pretreatment in surface reconstruction for sparing computer resources and improving reconstruction speed. However, existing methods often sacrifice the simplification precision to improve the simplification speed, or sacrifice the speed to improve precision. A proper balance between the simplification speed and the simplification accuracy is still a challenge. In this paper, we propose a new simplification method based on the importance of point. Named as detail feature points simplified algorithm (DFPSA), this algorithm has distinct processes to achieve improvements in three aspects. First, a rule of k neighborhood search is set to ensure the points found are the closest to the sample point. In this way, the accuracy of calculated normal vector of the point cloud is significantly improved, and the search speed is largely increased. Second, a formula that considers multiple characteristics for measuring the importance of point is proposed. Thereupon, the main detail features of the point cloud are preserved. Finally, an octree structure is employed to simplify the remaining points, through which holes in reconstructing point cloud are obviously reduced. The DFPSA is applied to four different data sets, and the corresponding results are compared with those of other five algorithms. The experimental results demonstrate that the DFPSA brings better simplification effects than existing counterparts, and the DFPSA not only can simplify point cloud but also has good effect in simplifying subject's narrow contours.INDEX TERMS 3D geometric point cloud simplification, detail feature point, k neighborhood search rule, the importance of point.

show abstract

“…In this paper, we use octree to simplify non-feature points. A simple schematic diagram of an octree structure [31] is shown in Fig. 1.…”

Section: Calculating the Normal Vectormentioning

confidence: 99%

A Novel Simplification Method for 3D Geometric Point Cloud Based on the Importance of Point

Liu

Fan

et al. 2019

IEEE Access

View full text Add to dashboard Cite

show abstract

“…PointNet is a neural network architecture that accepts raw point clouds as input and is robust with respect to input perturbation and corruption [ 24 ]. Taking into account the promising results of PointNet, several extensions have been developed [ 25 , 26 , 27 , 28 ].…”

Section: Methodsmentioning

confidence: 99%

Static Attitude Determination Using Convolutional Neural Networks

Santos

Seman

Bezerra

et al. 2021

Sensors

View full text Add to dashboard Cite

The need to estimate the orientation between frames of reference is crucial in spacecraft navigation. Robust algorithms for this type of problem have been built by following algebraic approaches, but data-driven solutions are becoming more appealing due to their stochastic nature. Hence, an approach based on convolutional neural networks in order to deal with measurement uncertainty in static attitude determination problems is proposed in this paper. PointNet models were trained with different datasets containing different numbers of observation vectors that were used to build attitude profile matrices, which were the inputs of the system. The uncertainty of measurements in the test scenarios was taken into consideration when choosing the best model. The proposed model, which used convolutional neural networks, proved to be less sensitive to higher noise than traditional algorithms, such as singular value decomposition (SVD), the q-method, the quaternion estimator (QUEST), and the second estimator of the optimal quaternion (ESOQ2).

show abstract

“…Deep learning is widely used in various fields, such as natural language processing [21], speech recognition [17], image processing [15], and point cloud processing [3,12,20,48,51], and others. Goodfellow et al [11].…”

Section: Related Study 21 3d Deep Learningmentioning

confidence: 99%

Semantic Segmentation for Full-Waveform LiDAR Data Using Local and Hierarchical Global Feature Extraction

Shinohara

Xiu

Matsuoka

2020

Proceedings of the 28th International Conference on Advances in Geographic Information Systems

View full text Add to dashboard Cite

During the last few years, in the field of computer vision, sophisticated deep learning methods have been developed to accomplish semantic segmentation tasks of 3D point cloud data. Additionally, many researchers have extended the applicability of these methods, such as PointNet or PointNet++, beyond semantic segmentation tasks of indoor scene data to large-scale outdoor scene data observed using airborne laser scanning systems equipped with light detection and ranging (LiDAR) technology. Most extant studies have only investigated geometric information (x, y, and z or longitude, latitude, and height) and have omitted rich radiometric information. Therefore, we aim to extend the applicability of deep learning-based model from the geometric data into radiometric data acquired with airborne full-waveform LiDAR without converting the waveform into 2D images or 3D voxels. We simultaneously train two models: a local module for local feature extraction and a global module for acquiring wide receptive fields for the waveform. Furthermore, our proposed model is based on waveform-aware convolutional techniques. We evaluate the effectiveness of the proposed method using benchmark large-scale outdoor scene data. By integrating the two outputs from the local module and the global module, our proposed model had achieved higher mean recall value 0.92 than previous methods and higher F1 scores for all six classes than the other 3D Deep Learning method. Therefore, our proposed network consisting of the local and global module successfully resolves the semantic segmentation task of full-waveform LiDAR data without requiring expert knowledge. CCS CONCEPTS • Computing methodologies → Scene understanding.

show abstract

Using Deep Learning in Semantic Classification for Point Cloud Data

Cited by 27 publications

References 25 publications

A Novel Simplification Method for 3D Geometric Point Cloud Based on the Importance of Point

A Novel Simplification Method for 3D Geometric Point Cloud Based on the Importance of Point

Static Attitude Determination Using Convolutional Neural Networks

Semantic Segmentation for Full-Waveform LiDAR Data Using Local and Hierarchical Global Feature Extraction

Contact Info

Product

Resources

About