A Synthetic Data Generation Approach With Dynamic Camera Poses for Long-Range Object Detection in AI Applications
Accurate long-range object detection is essential for applications such as security and surveillance. However, existing datasets often lack the complexity needed to represent real-world outdoor environments, resulting in limited performance of object detection algorithms at extended distances.Synthe...
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2024-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/10802893/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1850084984471158784 |
|---|---|
| author | Misbah Bibi Anam Nawaz Khan Muhammad Faseeh Qazi Waqas Khan Rashid Ahmad do-Hyeun Kim |
| author_facet | Misbah Bibi Anam Nawaz Khan Muhammad Faseeh Qazi Waqas Khan Rashid Ahmad do-Hyeun Kim |
| author_sort | Misbah Bibi |
| collection | DOAJ |
| description | Accurate long-range object detection is essential for applications such as security and surveillance. However, existing datasets often lack the complexity needed to represent real-world outdoor environments, resulting in limited performance of object detection algorithms at extended distances.Synthetic data generation offers a way to address these limitations by creating varied and realistic training scenarios. To address these limitations, we propose a novel approach utilizing BlenderProc procedural generation and photorealistic rendering to create a synthetic dataset that captures diverse and realistic outdoor scenes for the objects detection. We trained YOLO model on this dataset and compared its performance with standard YOLO model. Our approach achieved a precision of 94% and recall of 96% for detecting objects at distances exceeding 120 meters, demonstrating significant improvements over existing methods. These findings underscore the potential of advanced synthetic data generation techniques to enhance long-range object detection and address critical challenges in surveillance, remote sensing, and autonomous systems. |
| format | Article |
| id | doaj-art-1fc20c31eb9645ce876e5c2f0ebdad03 |
| institution | DOAJ |
| issn | 2169-3536 |
| language | English |
| publishDate | 2024-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj-art-1fc20c31eb9645ce876e5c2f0ebdad032025-08-20T02:43:50ZengIEEEIEEE Access2169-35362024-01-011219450519452010.1109/ACCESS.2024.351771710802893A Synthetic Data Generation Approach With Dynamic Camera Poses for Long-Range Object Detection in AI ApplicationsMisbah Bibi0https://orcid.org/0009-0006-0959-3852Anam Nawaz Khan1https://orcid.org/0000-0001-6260-5820Muhammad Faseeh2https://orcid.org/0009-0008-2375-7803Qazi Waqas Khan3https://orcid.org/0000-0002-4031-3920Rashid Ahmad4https://orcid.org/0000-0001-6922-7412do-Hyeun Kim5https://orcid.org/0000-0002-3457-2301Department of Computer Engineering, Jeju National University, Jeju-si, Republic of KoreaDepartment of Computer Engineering, Big Data Research Center, Jeju National University, Jeju-si, South KoreaDepartment of Computer Engineering, Jeju National University, Jeju-si, Republic of KoreaDepartment of Computer Engineering, Jeju National University, Jeju-si, Republic of KoreaFaculty of Computing and IT, Sohar University, Sohar, Sultanate of OmanDepartment of Computer Engineering, Big Data Research Center, Jeju National University, Jeju-si, South KoreaAccurate long-range object detection is essential for applications such as security and surveillance. However, existing datasets often lack the complexity needed to represent real-world outdoor environments, resulting in limited performance of object detection algorithms at extended distances.Synthetic data generation offers a way to address these limitations by creating varied and realistic training scenarios. To address these limitations, we propose a novel approach utilizing BlenderProc procedural generation and photorealistic rendering to create a synthetic dataset that captures diverse and realistic outdoor scenes for the objects detection. We trained YOLO model on this dataset and compared its performance with standard YOLO model. Our approach achieved a precision of 94% and recall of 96% for detecting objects at distances exceeding 120 meters, demonstrating significant improvements over existing methods. These findings underscore the potential of advanced synthetic data generation techniques to enhance long-range object detection and address critical challenges in surveillance, remote sensing, and autonomous systems.https://ieeexplore.ieee.org/document/10802893/Synthetic dataset generationphoto realistic renderingsecurity and surveillanceinstance segmentationlong distance object detection |
| spellingShingle | Misbah Bibi Anam Nawaz Khan Muhammad Faseeh Qazi Waqas Khan Rashid Ahmad do-Hyeun Kim A Synthetic Data Generation Approach With Dynamic Camera Poses for Long-Range Object Detection in AI Applications IEEE Access Synthetic dataset generation photo realistic rendering security and surveillance instance segmentation long distance object detection |
| title | A Synthetic Data Generation Approach With Dynamic Camera Poses for Long-Range Object Detection in AI Applications |
| title_full | A Synthetic Data Generation Approach With Dynamic Camera Poses for Long-Range Object Detection in AI Applications |
| title_fullStr | A Synthetic Data Generation Approach With Dynamic Camera Poses for Long-Range Object Detection in AI Applications |
| title_full_unstemmed | A Synthetic Data Generation Approach With Dynamic Camera Poses for Long-Range Object Detection in AI Applications |
| title_short | A Synthetic Data Generation Approach With Dynamic Camera Poses for Long-Range Object Detection in AI Applications |
| title_sort | synthetic data generation approach with dynamic camera poses for long range object detection in ai applications |
| topic | Synthetic dataset generation photo realistic rendering security and surveillance instance segmentation long distance object detection |
| url | https://ieeexplore.ieee.org/document/10802893/ |
| work_keys_str_mv | AT misbahbibi asyntheticdatagenerationapproachwithdynamiccameraposesforlongrangeobjectdetectioninaiapplications AT anamnawazkhan asyntheticdatagenerationapproachwithdynamiccameraposesforlongrangeobjectdetectioninaiapplications AT muhammadfaseeh asyntheticdatagenerationapproachwithdynamiccameraposesforlongrangeobjectdetectioninaiapplications AT qaziwaqaskhan asyntheticdatagenerationapproachwithdynamiccameraposesforlongrangeobjectdetectioninaiapplications AT rashidahmad asyntheticdatagenerationapproachwithdynamiccameraposesforlongrangeobjectdetectioninaiapplications AT dohyeunkim asyntheticdatagenerationapproachwithdynamiccameraposesforlongrangeobjectdetectioninaiapplications AT misbahbibi syntheticdatagenerationapproachwithdynamiccameraposesforlongrangeobjectdetectioninaiapplications AT anamnawazkhan syntheticdatagenerationapproachwithdynamiccameraposesforlongrangeobjectdetectioninaiapplications AT muhammadfaseeh syntheticdatagenerationapproachwithdynamiccameraposesforlongrangeobjectdetectioninaiapplications AT qaziwaqaskhan syntheticdatagenerationapproachwithdynamiccameraposesforlongrangeobjectdetectioninaiapplications AT rashidahmad syntheticdatagenerationapproachwithdynamiccameraposesforlongrangeobjectdetectioninaiapplications AT dohyeunkim syntheticdatagenerationapproachwithdynamiccameraposesforlongrangeobjectdetectioninaiapplications |