EmbeddedPigDet—Fast and Accurate Pig Detection for Embedded Board Implementations
Automated pig monitoring is an important issue in the surveillance environment of a pig farm. For a large-scale pig farm in particular, practical issues such as monitoring cost should be considered but such consideration based on low-cost embedded boards has not yet been reported. Since low-cost emb...
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2020-04-01
|
| Series: | Applied Sciences |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2076-3417/10/8/2878 |
| _version_ | 1827718188094193664 |
|---|---|
| author | Jihyun Seo Hanse Ahn Daewon Kim Sungju Lee Yongwha Chung Daihee Park |
| author_facet | Jihyun Seo Hanse Ahn Daewon Kim Sungju Lee Yongwha Chung Daihee Park |
| author_sort | Jihyun Seo |
| collection | DOAJ |
| description | Automated pig monitoring is an important issue in the surveillance environment of a pig farm. For a large-scale pig farm in particular, practical issues such as monitoring cost should be considered but such consideration based on low-cost embedded boards has not yet been reported. Since low-cost embedded boards have more limited computing power than typical PCs and have tradeoffs between execution speed and accuracy, achieving fast and accurate detection of individual pigs for “on-device” pig monitoring applications is very challenging. Therefore, in this paper, we propose a method for the fast detection of individual pigs by reducing the computational workload of 3 × 3 convolution in widely-used, deep learning-based object detectors. Then, in order to recover the accuracy of the “light-weight” deep learning-based object detector, we generate a three-channel composite image as its input image, through “simple” image preprocessing techniques. Our experimental results on an NVIDIA Jetson Nano embedded board show that the proposed method can improve the integrated performance of both execution speed and accuracy of widely-used, deep learning-based object detectors, by a factor of up to 8.7. |
| first_indexed | 2024-03-10T20:18:49Z |
| format | Article |
| id | doaj.art-e2582d0448ce4426aee26286560cc0c7 |
| institution | Directory Open Access Journal |
| issn | 2076-3417 |
| language | English |
| last_indexed | 2024-03-10T20:18:49Z |
| publishDate | 2020-04-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Applied Sciences |
| spelling | doaj.art-e2582d0448ce4426aee26286560cc0c72023-11-19T22:24:09ZengMDPI AGApplied Sciences2076-34172020-04-01108287810.3390/app10082878EmbeddedPigDet—Fast and Accurate Pig Detection for Embedded Board ImplementationsJihyun Seo0Hanse Ahn1Daewon Kim2Sungju Lee3Yongwha Chung4Daihee Park5Department of Computer Convergence Software, Korea University, Sejong 30019, KoreaDepartment of Computer Convergence Software, Korea University, Sejong 30019, KoreaDepartment of Computer Convergence Software, Korea University, Sejong 30019, KoreaDepartment of Software, Sangmyung University, Cheonan 31066, KoreaDepartment of Computer Convergence Software, Korea University, Sejong 30019, KoreaDepartment of Computer Convergence Software, Korea University, Sejong 30019, KoreaAutomated pig monitoring is an important issue in the surveillance environment of a pig farm. For a large-scale pig farm in particular, practical issues such as monitoring cost should be considered but such consideration based on low-cost embedded boards has not yet been reported. Since low-cost embedded boards have more limited computing power than typical PCs and have tradeoffs between execution speed and accuracy, achieving fast and accurate detection of individual pigs for “on-device” pig monitoring applications is very challenging. Therefore, in this paper, we propose a method for the fast detection of individual pigs by reducing the computational workload of 3 × 3 convolution in widely-used, deep learning-based object detectors. Then, in order to recover the accuracy of the “light-weight” deep learning-based object detector, we generate a three-channel composite image as its input image, through “simple” image preprocessing techniques. Our experimental results on an NVIDIA Jetson Nano embedded board show that the proposed method can improve the integrated performance of both execution speed and accuracy of widely-used, deep learning-based object detectors, by a factor of up to 8.7.https://www.mdpi.com/2076-3417/10/8/2878agriculture ITcomputer visionpig detectionembedded boardimage preprocessinglight-weight deep learning |
| spellingShingle | Jihyun Seo Hanse Ahn Daewon Kim Sungju Lee Yongwha Chung Daihee Park EmbeddedPigDet—Fast and Accurate Pig Detection for Embedded Board Implementations Applied Sciences agriculture IT computer vision pig detection embedded board image preprocessing light-weight deep learning |
| title | EmbeddedPigDet—Fast and Accurate Pig Detection for Embedded Board Implementations |
| title_full | EmbeddedPigDet—Fast and Accurate Pig Detection for Embedded Board Implementations |
| title_fullStr | EmbeddedPigDet—Fast and Accurate Pig Detection for Embedded Board Implementations |
| title_full_unstemmed | EmbeddedPigDet—Fast and Accurate Pig Detection for Embedded Board Implementations |
| title_short | EmbeddedPigDet—Fast and Accurate Pig Detection for Embedded Board Implementations |
| title_sort | embeddedpigdet fast and accurate pig detection for embedded board implementations |
| topic | agriculture IT computer vision pig detection embedded board image preprocessing light-weight deep learning |
| url | https://www.mdpi.com/2076-3417/10/8/2878 |
| work_keys_str_mv | AT jihyunseo embeddedpigdetfastandaccuratepigdetectionforembeddedboardimplementations AT hanseahn embeddedpigdetfastandaccuratepigdetectionforembeddedboardimplementations AT daewonkim embeddedpigdetfastandaccuratepigdetectionforembeddedboardimplementations AT sungjulee embeddedpigdetfastandaccuratepigdetectionforembeddedboardimplementations AT yongwhachung embeddedpigdetfastandaccuratepigdetectionforembeddedboardimplementations AT daiheepark embeddedpigdetfastandaccuratepigdetectionforembeddedboardimplementations |