LiDAR Sensor Parameter Augmentation and Data-Driven Influence Analysis on Deep-Learning-Based People Detection
Light detection and ranging (LiDAR) sensor technology for people detection offers a significant advantage in data protection. However, to design these systems cost- and energy-efficiently, the relationship between the measurement data and final object detection output with deep neural networks (DNNs...
Saved in:
| Main Authors: | , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2025-05-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1424-8220/25/10/3114 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1849326918859489280 |
|---|---|
| author | Lukas Haas Florian Sanne Johann Zedelmeier Subir Das Thomas Zeh Matthias Kuba Florian Bindges Martin Jakobi Alexander W. Koch |
| author_facet | Lukas Haas Florian Sanne Johann Zedelmeier Subir Das Thomas Zeh Matthias Kuba Florian Bindges Martin Jakobi Alexander W. Koch |
| author_sort | Lukas Haas |
| collection | DOAJ |
| description | Light detection and ranging (LiDAR) sensor technology for people detection offers a significant advantage in data protection. However, to design these systems cost- and energy-efficiently, the relationship between the measurement data and final object detection output with deep neural networks (DNNs) has to be elaborated. Therefore, this paper presents augmentation methods to analyze the influence of the distance, resolution, noise, and shading parameters of a LiDAR sensor in real point clouds for people detection. Furthermore, their influence on object detection using DNNs was investigated. A significant reduction in the quality requirements for the point clouds was possible for the measurement setup with only minor degradation on the object list level. The DNNs PointVoxel-Region-based Convolutional Neural Network (PV-RCNN) and Sparsely Embedded Convolutional Detection (SECOND) both only show a reduction in object detection of less than 5% with a reduced resolution of up to 32 factors, for an increase in distance of 4 factors, and with a Gaussian noise up to <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><mrow><mi>μ</mi><mo>=</mo><mn>0</mn></mrow></semantics></math></inline-formula> and <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><mrow><mi>σ</mi><mo>=</mo><mn>0.07</mn></mrow></semantics></math></inline-formula>. In addition, both networks require an unshaded height of approx. 0.5 m from a detected person’s head downwards to ensure good people detection performance without special training for these cases. The results obtained, such as shadowing information, are transferred to a software program to determine the minimum number of sensors and their orientation based on the mounting height of the sensor, the sensor parameters, and the ground area under consideration, both for detection at the point cloud level and object detection level. |
| format | Article |
| id | doaj-art-32701bd3758a49cf89266d7281b153aa |
| institution | Kabale University |
| issn | 1424-8220 |
| language | English |
| publishDate | 2025-05-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj-art-32701bd3758a49cf89266d7281b153aa2025-08-20T03:48:02ZengMDPI AGSensors1424-82202025-05-012510311410.3390/s25103114LiDAR Sensor Parameter Augmentation and Data-Driven Influence Analysis on Deep-Learning-Based People DetectionLukas Haas0Florian Sanne1Johann Zedelmeier2Subir Das3Thomas Zeh4Matthias Kuba5Florian Bindges6Martin Jakobi7Alexander W. Koch8IFM—Institute for Driver Assistance Systems and Connected Mobility, Kempten University of Applied Sciences, Junkerstraße 1A, 87734 Benningen, GermanyInstitute for Measurement Systems and Sensor Technology, Technical University of Munich, Theresienstr. 90, 80333 Munich, GermanyIFM—Institute for Driver Assistance Systems and Connected Mobility, Kempten University of Applied Sciences, Junkerstraße 1A, 87734 Benningen, GermanyIFM—Institute for Driver Assistance Systems and Connected Mobility, Kempten University of Applied Sciences, Junkerstraße 1A, 87734 Benningen, GermanyIFM—Institute for Driver Assistance Systems and Connected Mobility, Kempten University of Applied Sciences, Junkerstraße 1A, 87734 Benningen, GermanyFaculty of Electrical Engineering, Kempten University of Applied Sciences, Bahnhofstraße 61, 87435 Kempten, GermanyBlickfeld GmbH, Barthstr. 14, 80339 Munich, GermanyInstitute for Measurement Systems and Sensor Technology, Technical University of Munich, Theresienstr. 90, 80333 Munich, GermanyInstitute for Measurement Systems and Sensor Technology, Technical University of Munich, Theresienstr. 90, 80333 Munich, GermanyLight detection and ranging (LiDAR) sensor technology for people detection offers a significant advantage in data protection. However, to design these systems cost- and energy-efficiently, the relationship between the measurement data and final object detection output with deep neural networks (DNNs) has to be elaborated. Therefore, this paper presents augmentation methods to analyze the influence of the distance, resolution, noise, and shading parameters of a LiDAR sensor in real point clouds for people detection. Furthermore, their influence on object detection using DNNs was investigated. A significant reduction in the quality requirements for the point clouds was possible for the measurement setup with only minor degradation on the object list level. The DNNs PointVoxel-Region-based Convolutional Neural Network (PV-RCNN) and Sparsely Embedded Convolutional Detection (SECOND) both only show a reduction in object detection of less than 5% with a reduced resolution of up to 32 factors, for an increase in distance of 4 factors, and with a Gaussian noise up to <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><mrow><mi>μ</mi><mo>=</mo><mn>0</mn></mrow></semantics></math></inline-formula> and <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><mrow><mi>σ</mi><mo>=</mo><mn>0.07</mn></mrow></semantics></math></inline-formula>. In addition, both networks require an unshaded height of approx. 0.5 m from a detected person’s head downwards to ensure good people detection performance without special training for these cases. The results obtained, such as shadowing information, are transferred to a software program to determine the minimum number of sensors and their orientation based on the mounting height of the sensor, the sensor parameters, and the ground area under consideration, both for detection at the point cloud level and object detection level.https://www.mdpi.com/1424-8220/25/10/3114deep learningLiDAR sensorneural networkspeople detectionpoint cloudpoint cloud augmentation |
| spellingShingle | Lukas Haas Florian Sanne Johann Zedelmeier Subir Das Thomas Zeh Matthias Kuba Florian Bindges Martin Jakobi Alexander W. Koch LiDAR Sensor Parameter Augmentation and Data-Driven Influence Analysis on Deep-Learning-Based People Detection Sensors deep learning LiDAR sensor neural networks people detection point cloud point cloud augmentation |
| title | LiDAR Sensor Parameter Augmentation and Data-Driven Influence Analysis on Deep-Learning-Based People Detection |
| title_full | LiDAR Sensor Parameter Augmentation and Data-Driven Influence Analysis on Deep-Learning-Based People Detection |
| title_fullStr | LiDAR Sensor Parameter Augmentation and Data-Driven Influence Analysis on Deep-Learning-Based People Detection |
| title_full_unstemmed | LiDAR Sensor Parameter Augmentation and Data-Driven Influence Analysis on Deep-Learning-Based People Detection |
| title_short | LiDAR Sensor Parameter Augmentation and Data-Driven Influence Analysis on Deep-Learning-Based People Detection |
| title_sort | lidar sensor parameter augmentation and data driven influence analysis on deep learning based people detection |
| topic | deep learning LiDAR sensor neural networks people detection point cloud point cloud augmentation |
| url | https://www.mdpi.com/1424-8220/25/10/3114 |
| work_keys_str_mv | AT lukashaas lidarsensorparameteraugmentationanddatadriveninfluenceanalysisondeeplearningbasedpeopledetection AT floriansanne lidarsensorparameteraugmentationanddatadriveninfluenceanalysisondeeplearningbasedpeopledetection AT johannzedelmeier lidarsensorparameteraugmentationanddatadriveninfluenceanalysisondeeplearningbasedpeopledetection AT subirdas lidarsensorparameteraugmentationanddatadriveninfluenceanalysisondeeplearningbasedpeopledetection AT thomaszeh lidarsensorparameteraugmentationanddatadriveninfluenceanalysisondeeplearningbasedpeopledetection AT matthiaskuba lidarsensorparameteraugmentationanddatadriveninfluenceanalysisondeeplearningbasedpeopledetection AT florianbindges lidarsensorparameteraugmentationanddatadriveninfluenceanalysisondeeplearningbasedpeopledetection AT martinjakobi lidarsensorparameteraugmentationanddatadriveninfluenceanalysisondeeplearningbasedpeopledetection AT alexanderwkoch lidarsensorparameteraugmentationanddatadriveninfluenceanalysisondeeplearningbasedpeopledetection |