Sticky Trap-Embedded Machine Vision for Tea Pest Monitoring: A Cross-Domain Transfer Learning Framework Addressing Few-Shot Small Target Detection
Pest infestations have always been a major factor affecting tea production. Real-time detection of tea pests using machine vision is a mainstream method in modern agricultural pest control. Currently, there is a notable absence of machine vision devices capable of real-time monitoring for small-size...
Saved in:
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2025-03-01
|
| Series: | Agronomy |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2073-4395/15/3/693 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1849392754100011008 |
|---|---|
| author | Kunhong Li Yi Li Xuan Wen Jingsha Shi Linsi Yang Yuyang Xiao Xiaosong Lu Jiong Mu |
| author_facet | Kunhong Li Yi Li Xuan Wen Jingsha Shi Linsi Yang Yuyang Xiao Xiaosong Lu Jiong Mu |
| author_sort | Kunhong Li |
| collection | DOAJ |
| description | Pest infestations have always been a major factor affecting tea production. Real-time detection of tea pests using machine vision is a mainstream method in modern agricultural pest control. Currently, there is a notable absence of machine vision devices capable of real-time monitoring for small-sized tea pests in the market, and the scarcity of open-source datasets available for tea pest detection remains a critical limitation. This manuscript proposes a YOLOv8-FasterTea pest detection algorithm based on cross-domain transfer learning, which was successfully deployed in a novel tea pest monitoring device. The proposed method leverages transfer learning from the natural language character domain to the tea pest detection domain, termed cross-domain transfer learning, which is based on the complex and small characteristics shared by natural language characters and tea pests. With sufficient samples in the language character domain, transfer learning can effectively enhance the tiny and complex feature extraction capabilities of deep networks in the pest domain and mitigate the few-shot learning problem in tea pest detection. The information and texture features of small tea pests are more likely to be lost with the layers of a neural network becoming deep. Therefore, the proposed method, YOLOv8-FasterTea, removes the P5 layer and adds a P2 small target detection layer based on the YOLOv8 model. Additionally, the original C2f module is replaced with lighter convolutional modules to reduce the loss of information about small target pests. Finally, this manuscript successfully applies the algorithm to outdoor pest monitoring equipment. Experimental results demonstrate that, on a small sample yellow board pest dataset, the mAP@.5 value of the model increased by approximately 6%, on average, after transfer learning. The YOLOv8-FasterTea model improved the mAP@.5 value by 3.7%, while the model size was reduced by 46.6%. |
| format | Article |
| id | doaj-art-1a4d0fe2cd4947cf8439cd5dcbfcc450 |
| institution | Kabale University |
| issn | 2073-4395 |
| language | English |
| publishDate | 2025-03-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Agronomy |
| spelling | doaj-art-1a4d0fe2cd4947cf8439cd5dcbfcc4502025-08-20T03:40:42ZengMDPI AGAgronomy2073-43952025-03-0115369310.3390/agronomy15030693Sticky Trap-Embedded Machine Vision for Tea Pest Monitoring: A Cross-Domain Transfer Learning Framework Addressing Few-Shot Small Target DetectionKunhong Li0Yi Li1Xuan Wen2Jingsha Shi3Linsi Yang4Yuyang Xiao5Xiaosong Lu6Jiong Mu7College of Information Engineering, Sichuan Agricultural University, Ya’an 625000, ChinaCollege of Information Engineering, Sichuan Agricultural University, Ya’an 625000, ChinaComputer Science and Technology, Henan Institute of Science and Technology, Xinxiang 453700, ChinaCollege of Information Engineering, Sichuan Agricultural University, Ya’an 625000, ChinaCollege of Information Engineering, Sichuan Agricultural University, Ya’an 625000, ChinaCollege of Information Engineering, Sichuan Agricultural University, Ya’an 625000, ChinaYa’an Guangju Agricultural Development Co., Ltd., Ya’an 625100, ChinaCollege of Information Engineering, Sichuan Agricultural University, Ya’an 625000, ChinaPest infestations have always been a major factor affecting tea production. Real-time detection of tea pests using machine vision is a mainstream method in modern agricultural pest control. Currently, there is a notable absence of machine vision devices capable of real-time monitoring for small-sized tea pests in the market, and the scarcity of open-source datasets available for tea pest detection remains a critical limitation. This manuscript proposes a YOLOv8-FasterTea pest detection algorithm based on cross-domain transfer learning, which was successfully deployed in a novel tea pest monitoring device. The proposed method leverages transfer learning from the natural language character domain to the tea pest detection domain, termed cross-domain transfer learning, which is based on the complex and small characteristics shared by natural language characters and tea pests. With sufficient samples in the language character domain, transfer learning can effectively enhance the tiny and complex feature extraction capabilities of deep networks in the pest domain and mitigate the few-shot learning problem in tea pest detection. The information and texture features of small tea pests are more likely to be lost with the layers of a neural network becoming deep. Therefore, the proposed method, YOLOv8-FasterTea, removes the P5 layer and adds a P2 small target detection layer based on the YOLOv8 model. Additionally, the original C2f module is replaced with lighter convolutional modules to reduce the loss of information about small target pests. Finally, this manuscript successfully applies the algorithm to outdoor pest monitoring equipment. Experimental results demonstrate that, on a small sample yellow board pest dataset, the mAP@.5 value of the model increased by approximately 6%, on average, after transfer learning. The YOLOv8-FasterTea model improved the mAP@.5 value by 3.7%, while the model size was reduced by 46.6%.https://www.mdpi.com/2073-4395/15/3/693machine visionpest detectionsmall targettransfer learning |
| spellingShingle | Kunhong Li Yi Li Xuan Wen Jingsha Shi Linsi Yang Yuyang Xiao Xiaosong Lu Jiong Mu Sticky Trap-Embedded Machine Vision for Tea Pest Monitoring: A Cross-Domain Transfer Learning Framework Addressing Few-Shot Small Target Detection Agronomy machine vision pest detection small target transfer learning |
| title | Sticky Trap-Embedded Machine Vision for Tea Pest Monitoring: A Cross-Domain Transfer Learning Framework Addressing Few-Shot Small Target Detection |
| title_full | Sticky Trap-Embedded Machine Vision for Tea Pest Monitoring: A Cross-Domain Transfer Learning Framework Addressing Few-Shot Small Target Detection |
| title_fullStr | Sticky Trap-Embedded Machine Vision for Tea Pest Monitoring: A Cross-Domain Transfer Learning Framework Addressing Few-Shot Small Target Detection |
| title_full_unstemmed | Sticky Trap-Embedded Machine Vision for Tea Pest Monitoring: A Cross-Domain Transfer Learning Framework Addressing Few-Shot Small Target Detection |
| title_short | Sticky Trap-Embedded Machine Vision for Tea Pest Monitoring: A Cross-Domain Transfer Learning Framework Addressing Few-Shot Small Target Detection |
| title_sort | sticky trap embedded machine vision for tea pest monitoring a cross domain transfer learning framework addressing few shot small target detection |
| topic | machine vision pest detection small target transfer learning |
| url | https://www.mdpi.com/2073-4395/15/3/693 |
| work_keys_str_mv | AT kunhongli stickytrapembeddedmachinevisionforteapestmonitoringacrossdomaintransferlearningframeworkaddressingfewshotsmalltargetdetection AT yili stickytrapembeddedmachinevisionforteapestmonitoringacrossdomaintransferlearningframeworkaddressingfewshotsmalltargetdetection AT xuanwen stickytrapembeddedmachinevisionforteapestmonitoringacrossdomaintransferlearningframeworkaddressingfewshotsmalltargetdetection AT jingshashi stickytrapembeddedmachinevisionforteapestmonitoringacrossdomaintransferlearningframeworkaddressingfewshotsmalltargetdetection AT linsiyang stickytrapembeddedmachinevisionforteapestmonitoringacrossdomaintransferlearningframeworkaddressingfewshotsmalltargetdetection AT yuyangxiao stickytrapembeddedmachinevisionforteapestmonitoringacrossdomaintransferlearningframeworkaddressingfewshotsmalltargetdetection AT xiaosonglu stickytrapembeddedmachinevisionforteapestmonitoringacrossdomaintransferlearningframeworkaddressingfewshotsmalltargetdetection AT jiongmu stickytrapembeddedmachinevisionforteapestmonitoringacrossdomaintransferlearningframeworkaddressingfewshotsmalltargetdetection |