Real-Time Image Detection for Edge Devices: A Peach Fruit Detection Application
Within the scope of precision agriculture, many applications have been developed to support decision making and yield enhancement. Fruit detection has attracted considerable attention from researchers, and it can be used offline. In contrast, some applications, such as robot vision in orchards, requ...
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2022-11-01
|
| Series: | Future Internet |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1999-5903/14/11/323 |
| _version_ | 1827646498540617728 |
|---|---|
| author | Eduardo Assunção Pedro D. Gaspar Khadijeh Alibabaei Maria P. Simões Hugo Proença Vasco N. G. J. Soares João M. L. P. Caldeira |
| author_facet | Eduardo Assunção Pedro D. Gaspar Khadijeh Alibabaei Maria P. Simões Hugo Proença Vasco N. G. J. Soares João M. L. P. Caldeira |
| author_sort | Eduardo Assunção |
| collection | DOAJ |
| description | Within the scope of precision agriculture, many applications have been developed to support decision making and yield enhancement. Fruit detection has attracted considerable attention from researchers, and it can be used offline. In contrast, some applications, such as robot vision in orchards, require computer vision models to run on edge devices while performing inferences at high speed. In this area, most modern applications use an integrated graphics processing unit (GPU). In this work, we propose the use of a tensor processing unit (TPU) accelerator with a Raspberry Pi target device and the state-of-the-art, lightweight, and hardware-aware MobileDet detector model. Our contribution is the extension of the possibilities of using accelerators (the TPU) for edge devices in precision agriculture. The proposed method was evaluated using a novel dataset of peaches with three cultivars, which will be made available for further studies. The model achieved an average precision (AP) of 88.2% and a performance of 19.84 frames per second (FPS) at an image size of 640 × 480. The results obtained show that the TPU accelerator can be an excellent alternative for processing on the edge in precision agriculture. |
| first_indexed | 2024-03-09T19:03:29Z |
| format | Article |
| id | doaj.art-61a8dacc8f89409b9e62227ba2da96bd |
| institution | Directory Open Access Journal |
| issn | 1999-5903 |
| language | English |
| last_indexed | 2024-03-09T19:03:29Z |
| publishDate | 2022-11-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Future Internet |
| spelling | doaj.art-61a8dacc8f89409b9e62227ba2da96bd2023-11-24T04:46:01ZengMDPI AGFuture Internet1999-59032022-11-01141132310.3390/fi14110323Real-Time Image Detection for Edge Devices: A Peach Fruit Detection ApplicationEduardo Assunção0Pedro D. Gaspar1Khadijeh Alibabaei2Maria P. Simões3Hugo Proença4Vasco N. G. J. Soares5João M. L. P. Caldeira6C-MAST Center for Mechanical and Aerospace Science and Technologies, University of Beira Interior, 6201-001 Covilhã, PortugalC-MAST Center for Mechanical and Aerospace Science and Technologies, University of Beira Interior, 6201-001 Covilhã, PortugalC-MAST Center for Mechanical and Aerospace Science and Technologies, University of Beira Interior, 6201-001 Covilhã, PortugalPolytechnic Institute of Castelo Branco, Av. Pedro Álvares Cabral nº 12, 6000-084 Castelo Branco, PortugalInstituto de Telecomunicações, University of Beira Interior, Rua Marquês d’Ávila e Bolama, 6201-001 Covilhã, PortugalInstituto de Telecomunicações, University of Beira Interior, Rua Marquês d’Ávila e Bolama, 6201-001 Covilhã, PortugalInstituto de Telecomunicações, University of Beira Interior, Rua Marquês d’Ávila e Bolama, 6201-001 Covilhã, PortugalWithin the scope of precision agriculture, many applications have been developed to support decision making and yield enhancement. Fruit detection has attracted considerable attention from researchers, and it can be used offline. In contrast, some applications, such as robot vision in orchards, require computer vision models to run on edge devices while performing inferences at high speed. In this area, most modern applications use an integrated graphics processing unit (GPU). In this work, we propose the use of a tensor processing unit (TPU) accelerator with a Raspberry Pi target device and the state-of-the-art, lightweight, and hardware-aware MobileDet detector model. Our contribution is the extension of the possibilities of using accelerators (the TPU) for edge devices in precision agriculture. The proposed method was evaluated using a novel dataset of peaches with three cultivars, which will be made available for further studies. The model achieved an average precision (AP) of 88.2% and a performance of 19.84 frames per second (FPS) at an image size of 640 × 480. The results obtained show that the TPU accelerator can be an excellent alternative for processing on the edge in precision agriculture.https://www.mdpi.com/1999-5903/14/11/323deep learningedge deviceobject detectionprecision agricultureTPU accelerator |
| spellingShingle | Eduardo Assunção Pedro D. Gaspar Khadijeh Alibabaei Maria P. Simões Hugo Proença Vasco N. G. J. Soares João M. L. P. Caldeira Real-Time Image Detection for Edge Devices: A Peach Fruit Detection Application Future Internet deep learning edge device object detection precision agriculture TPU accelerator |
| title | Real-Time Image Detection for Edge Devices: A Peach Fruit Detection Application |
| title_full | Real-Time Image Detection for Edge Devices: A Peach Fruit Detection Application |
| title_fullStr | Real-Time Image Detection for Edge Devices: A Peach Fruit Detection Application |
| title_full_unstemmed | Real-Time Image Detection for Edge Devices: A Peach Fruit Detection Application |
| title_short | Real-Time Image Detection for Edge Devices: A Peach Fruit Detection Application |
| title_sort | real time image detection for edge devices a peach fruit detection application |
| topic | deep learning edge device object detection precision agriculture TPU accelerator |
| url | https://www.mdpi.com/1999-5903/14/11/323 |
| work_keys_str_mv | AT eduardoassuncao realtimeimagedetectionforedgedevicesapeachfruitdetectionapplication AT pedrodgaspar realtimeimagedetectionforedgedevicesapeachfruitdetectionapplication AT khadijehalibabaei realtimeimagedetectionforedgedevicesapeachfruitdetectionapplication AT mariapsimoes realtimeimagedetectionforedgedevicesapeachfruitdetectionapplication AT hugoproenca realtimeimagedetectionforedgedevicesapeachfruitdetectionapplication AT vascongjsoares realtimeimagedetectionforedgedevicesapeachfruitdetectionapplication AT joaomlpcaldeira realtimeimagedetectionforedgedevicesapeachfruitdetectionapplication |