Incorporating Handcrafted Features into Deep Learning for Point Cloud Classification
Point cloud classification is an important task in point cloud data analysis. Traditional point cloud classification is conducted primarily on the basis of specific handcrafted features with a specific classifier and is often capable of producing satisfactory results. However, the extraction of cruc...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2020-11-01
|
| Series: | Remote Sensing |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2072-4292/12/22/3713 |
| _version_ | 1797548098303557632 |
|---|---|
| author | Pai-Hui Hsu Zong-Yi Zhuang |
| author_facet | Pai-Hui Hsu Zong-Yi Zhuang |
| author_sort | Pai-Hui Hsu |
| collection | DOAJ |
| description | Point cloud classification is an important task in point cloud data analysis. Traditional point cloud classification is conducted primarily on the basis of specific handcrafted features with a specific classifier and is often capable of producing satisfactory results. However, the extraction of crucial handcrafted features hinges on sufficient knowledge of the field and substantial experience. In contrast, while powerful deep learning algorithms possess the ability to learn features automatically, it normally requires complex network architecture and a considerable amount of calculation time to attain better accuracy of classification. In order to combine the advantages of both the methods, in this study, we integrated the handcrafted features, whose benefits were confirmed by previous studies, into a deep learning network, in the hopes of solving the problem of insufficient extraction of specific features and enabling the network to recognise other effective features through automatic learning. This was done to achieve the performance of a complex model by using a simple model and fulfil the application requirements of the remote sensing domain. As indicated by the experimental results, the integration of handcrafted features into the simple and fast-calculating PointNet model could generate a classification result that bore comparison with that generated by a complex network model such as PointNet++ or KPConv. |
| first_indexed | 2024-03-10T14:53:38Z |
| format | Article |
| id | doaj.art-21c9397da4744929bd93e6af518bdee3 |
| institution | Directory Open Access Journal |
| issn | 2072-4292 |
| language | English |
| last_indexed | 2024-03-10T14:53:38Z |
| publishDate | 2020-11-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Remote Sensing |
| spelling | doaj.art-21c9397da4744929bd93e6af518bdee32023-11-20T20:43:20ZengMDPI AGRemote Sensing2072-42922020-11-011222371310.3390/rs12223713Incorporating Handcrafted Features into Deep Learning for Point Cloud ClassificationPai-Hui Hsu0Zong-Yi Zhuang1Department of Civil Engineering, National Taiwan University, No.1, Sec.4, Roosevelt Rd. Taipei City 10617, TaiwanDepartment of Civil Engineering, National Taiwan University, No.1, Sec.4, Roosevelt Rd. Taipei City 10617, TaiwanPoint cloud classification is an important task in point cloud data analysis. Traditional point cloud classification is conducted primarily on the basis of specific handcrafted features with a specific classifier and is often capable of producing satisfactory results. However, the extraction of crucial handcrafted features hinges on sufficient knowledge of the field and substantial experience. In contrast, while powerful deep learning algorithms possess the ability to learn features automatically, it normally requires complex network architecture and a considerable amount of calculation time to attain better accuracy of classification. In order to combine the advantages of both the methods, in this study, we integrated the handcrafted features, whose benefits were confirmed by previous studies, into a deep learning network, in the hopes of solving the problem of insufficient extraction of specific features and enabling the network to recognise other effective features through automatic learning. This was done to achieve the performance of a complex model by using a simple model and fulfil the application requirements of the remote sensing domain. As indicated by the experimental results, the integration of handcrafted features into the simple and fast-calculating PointNet model could generate a classification result that bore comparison with that generated by a complex network model such as PointNet++ or KPConv.https://www.mdpi.com/2072-4292/12/22/3713point cloudfeature extractionclassificationdeep learning |
| spellingShingle | Pai-Hui Hsu Zong-Yi Zhuang Incorporating Handcrafted Features into Deep Learning for Point Cloud Classification Remote Sensing point cloud feature extraction classification deep learning |
| title | Incorporating Handcrafted Features into Deep Learning for Point Cloud Classification |
| title_full | Incorporating Handcrafted Features into Deep Learning for Point Cloud Classification |
| title_fullStr | Incorporating Handcrafted Features into Deep Learning for Point Cloud Classification |
| title_full_unstemmed | Incorporating Handcrafted Features into Deep Learning for Point Cloud Classification |
| title_short | Incorporating Handcrafted Features into Deep Learning for Point Cloud Classification |
| title_sort | incorporating handcrafted features into deep learning for point cloud classification |
| topic | point cloud feature extraction classification deep learning |
| url | https://www.mdpi.com/2072-4292/12/22/3713 |
| work_keys_str_mv | AT paihuihsu incorporatinghandcraftedfeaturesintodeeplearningforpointcloudclassification AT zongyizhuang incorporatinghandcraftedfeaturesintodeeplearningforpointcloudclassification |