Online Self-Calibration of 3D Measurement Sensors Using a Voxel-Based Network
Multi-sensor fusion is important in the field of autonomous driving. A basic prerequisite for multi-sensor fusion is calibration between sensors. Such calibrations must be accurate and need to be performed online. Traditional calibration methods have strict rules. In contrast, the latest online cali...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2022-08-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1424-8220/22/17/6447 |
| _version_ | 1797493192509095936 |
|---|---|
| author | Jingyu Song Joonwoong Lee |
| author_facet | Jingyu Song Joonwoong Lee |
| author_sort | Jingyu Song |
| collection | DOAJ |
| description | Multi-sensor fusion is important in the field of autonomous driving. A basic prerequisite for multi-sensor fusion is calibration between sensors. Such calibrations must be accurate and need to be performed online. Traditional calibration methods have strict rules. In contrast, the latest online calibration methods based on convolutional neural networks (CNNs) have gone beyond the limits of the conventional methods. We propose a novel algorithm for online self-calibration between sensors using voxels and three-dimensional (3D) convolution kernels. The proposed approach has the following features: (1) it is intended for calibration between sensors that measure 3D space; (2) the proposed network is capable of end-to-end learning; (3) the input 3D point cloud is converted to voxel information; (4) it uses five networks that process voxel information, and it improves calibration accuracy through iterative refinement of the output of the five networks and temporal filtering. We use the KITTI and Oxford datasets to evaluate the calibration performance of the proposed method. The proposed method achieves a rotation error of less than 0.1° and a translation error of less than 1 cm on both the KITTI and Oxford datasets. |
| first_indexed | 2024-03-10T01:16:29Z |
| format | Article |
| id | doaj.art-c9b44670e1eb4828b069d8404ef3d23c |
| institution | Directory Open Access Journal |
| issn | 1424-8220 |
| language | English |
| last_indexed | 2024-03-10T01:16:29Z |
| publishDate | 2022-08-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj.art-c9b44670e1eb4828b069d8404ef3d23c2023-11-23T14:08:39ZengMDPI AGSensors1424-82202022-08-012217644710.3390/s22176447Online Self-Calibration of 3D Measurement Sensors Using a Voxel-Based NetworkJingyu Song0Joonwoong Lee1Department of Industrial Engineering, Chonnam National University, 77 Yongbong-ro, Buk-gu, Gwangju 61186, KoreaDepartment of Industrial Engineering, Chonnam National University, 77 Yongbong-ro, Buk-gu, Gwangju 61186, KoreaMulti-sensor fusion is important in the field of autonomous driving. A basic prerequisite for multi-sensor fusion is calibration between sensors. Such calibrations must be accurate and need to be performed online. Traditional calibration methods have strict rules. In contrast, the latest online calibration methods based on convolutional neural networks (CNNs) have gone beyond the limits of the conventional methods. We propose a novel algorithm for online self-calibration between sensors using voxels and three-dimensional (3D) convolution kernels. The proposed approach has the following features: (1) it is intended for calibration between sensors that measure 3D space; (2) the proposed network is capable of end-to-end learning; (3) the input 3D point cloud is converted to voxel information; (4) it uses five networks that process voxel information, and it improves calibration accuracy through iterative refinement of the output of the five networks and temporal filtering. We use the KITTI and Oxford datasets to evaluate the calibration performance of the proposed method. The proposed method achieves a rotation error of less than 0.1° and a translation error of less than 1 cm on both the KITTI and Oxford datasets.https://www.mdpi.com/1424-8220/22/17/6447online self-calibrationconvolutional neural networkvoxel information |
| spellingShingle | Jingyu Song Joonwoong Lee Online Self-Calibration of 3D Measurement Sensors Using a Voxel-Based Network Sensors online self-calibration convolutional neural network voxel information |
| title | Online Self-Calibration of 3D Measurement Sensors Using a Voxel-Based Network |
| title_full | Online Self-Calibration of 3D Measurement Sensors Using a Voxel-Based Network |
| title_fullStr | Online Self-Calibration of 3D Measurement Sensors Using a Voxel-Based Network |
| title_full_unstemmed | Online Self-Calibration of 3D Measurement Sensors Using a Voxel-Based Network |
| title_short | Online Self-Calibration of 3D Measurement Sensors Using a Voxel-Based Network |
| title_sort | online self calibration of 3d measurement sensors using a voxel based network |
| topic | online self-calibration convolutional neural network voxel information |
| url | https://www.mdpi.com/1424-8220/22/17/6447 |
| work_keys_str_mv | AT jingyusong onlineselfcalibrationof3dmeasurementsensorsusingavoxelbasednetwork AT joonwoonglee onlineselfcalibrationof3dmeasurementsensorsusingavoxelbasednetwork |