A fast and efficient green apple object detection model based on Foveabox
Fruit object detection is crucial for automatic harvesting systems, serving applications such as orchard yield measurement and fruit harvesting. In order to achieve fast recognition and localization of green apples and meet the real-time working requirements of the vision system of harvesting robots...
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Springer
2022-09-01
|
| Series: | Journal of King Saud University: Computer and Information Sciences |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S1319157822000179 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1849315846987448320 |
|---|---|
| author | Weikuan Jia Zhifen Wang Zhonghua Zhang Xinbo Yang Sujuan Hou Yuanjie Zheng |
| author_facet | Weikuan Jia Zhifen Wang Zhonghua Zhang Xinbo Yang Sujuan Hou Yuanjie Zheng |
| author_sort | Weikuan Jia |
| collection | DOAJ |
| description | Fruit object detection is crucial for automatic harvesting systems, serving applications such as orchard yield measurement and fruit harvesting. In order to achieve fast recognition and localization of green apples and meet the real-time working requirements of the vision system of harvesting robots, a fast optimized Foveabox detection model (Fast-FDM) is proposed. Fast-FDM uses an optimized form of anchor-free Foveabox to accurately and efficiently detect green apples in harvesting environments. Specifically, the EfficientNetV2-S with fast training and small size is used as the backbone network, a weighted bi-directional feature pyramid network (BiFPN) is employed as the feature extraction network to fuse multi-scale features easily and fast, and then the fused features are fed to the fovea head prediction network for the classification and bounding box prediction. Furthermore, an adaptive training sample selection (ATSS) method is adopted to directly select positive and negative samples, allowing green fruits of different scales to obtain higher recall and achieve more accurate green apple detection. Experimental results show that the proposed Fast-FDM realizes a mean average precision (mAP) of 62.3% for green apple detection using fewer parameters and floating point of operations (FLOPs), achieving better trade-offs between accuracy and detection efficiency. |
| format | Article |
| id | doaj-art-97aa9b65c65b43a191a0b8a8b110f462 |
| institution | Kabale University |
| issn | 1319-1578 |
| language | English |
| publishDate | 2022-09-01 |
| publisher | Springer |
| record_format | Article |
| series | Journal of King Saud University: Computer and Information Sciences |
| spelling | doaj-art-97aa9b65c65b43a191a0b8a8b110f4622025-08-20T03:52:02ZengSpringerJournal of King Saud University: Computer and Information Sciences1319-15782022-09-013485156516910.1016/j.jksuci.2022.01.005A fast and efficient green apple object detection model based on FoveaboxWeikuan Jia0Zhifen Wang1Zhonghua Zhang2Xinbo Yang3Sujuan Hou4Yuanjie Zheng5School of Information Science and Engineering, Shandong Normal University, Jinan 250358, China; Key Laboratory of Facility Agriculture Measurement and Control Technology and Equipment of Machinery Industry, Zhenjiang 212013, China; Corresponding authors at: School of Information Science and Engineering, Shandong Normal University, Jinan 250358, China (W. Jia, S. Hou).School of Information Science and Engineering, Shandong Normal University, Jinan 250358, ChinaSchool of Information Science and Engineering, Shandong Normal University, Jinan 250358, ChinaSchool of Information Science and Engineering, Shandong Normal University, Jinan 250358, ChinaSchool of Information Science and Engineering, Shandong Normal University, Jinan 250358, China; Corresponding authors at: School of Information Science and Engineering, Shandong Normal University, Jinan 250358, China (W. Jia, S. Hou).School of Information Science and Engineering, Shandong Normal University, Jinan 250358, ChinaFruit object detection is crucial for automatic harvesting systems, serving applications such as orchard yield measurement and fruit harvesting. In order to achieve fast recognition and localization of green apples and meet the real-time working requirements of the vision system of harvesting robots, a fast optimized Foveabox detection model (Fast-FDM) is proposed. Fast-FDM uses an optimized form of anchor-free Foveabox to accurately and efficiently detect green apples in harvesting environments. Specifically, the EfficientNetV2-S with fast training and small size is used as the backbone network, a weighted bi-directional feature pyramid network (BiFPN) is employed as the feature extraction network to fuse multi-scale features easily and fast, and then the fused features are fed to the fovea head prediction network for the classification and bounding box prediction. Furthermore, an adaptive training sample selection (ATSS) method is adopted to directly select positive and negative samples, allowing green fruits of different scales to obtain higher recall and achieve more accurate green apple detection. Experimental results show that the proposed Fast-FDM realizes a mean average precision (mAP) of 62.3% for green apple detection using fewer parameters and floating point of operations (FLOPs), achieving better trade-offs between accuracy and detection efficiency.http://www.sciencedirect.com/science/article/pii/S1319157822000179Fast-FDMObject detectionGreen appleFoveaBoxATSS |
| spellingShingle | Weikuan Jia Zhifen Wang Zhonghua Zhang Xinbo Yang Sujuan Hou Yuanjie Zheng A fast and efficient green apple object detection model based on Foveabox Journal of King Saud University: Computer and Information Sciences Fast-FDM Object detection Green apple FoveaBox ATSS |
| title | A fast and efficient green apple object detection model based on Foveabox |
| title_full | A fast and efficient green apple object detection model based on Foveabox |
| title_fullStr | A fast and efficient green apple object detection model based on Foveabox |
| title_full_unstemmed | A fast and efficient green apple object detection model based on Foveabox |
| title_short | A fast and efficient green apple object detection model based on Foveabox |
| title_sort | fast and efficient green apple object detection model based on foveabox |
| topic | Fast-FDM Object detection Green apple FoveaBox ATSS |
| url | http://www.sciencedirect.com/science/article/pii/S1319157822000179 |
| work_keys_str_mv | AT weikuanjia afastandefficientgreenappleobjectdetectionmodelbasedonfoveabox AT zhifenwang afastandefficientgreenappleobjectdetectionmodelbasedonfoveabox AT zhonghuazhang afastandefficientgreenappleobjectdetectionmodelbasedonfoveabox AT xinboyang afastandefficientgreenappleobjectdetectionmodelbasedonfoveabox AT sujuanhou afastandefficientgreenappleobjectdetectionmodelbasedonfoveabox AT yuanjiezheng afastandefficientgreenappleobjectdetectionmodelbasedonfoveabox AT weikuanjia fastandefficientgreenappleobjectdetectionmodelbasedonfoveabox AT zhifenwang fastandefficientgreenappleobjectdetectionmodelbasedonfoveabox AT zhonghuazhang fastandefficientgreenappleobjectdetectionmodelbasedonfoveabox AT xinboyang fastandefficientgreenappleobjectdetectionmodelbasedonfoveabox AT sujuanhou fastandefficientgreenappleobjectdetectionmodelbasedonfoveabox AT yuanjiezheng fastandefficientgreenappleobjectdetectionmodelbasedonfoveabox |