Different model detection results comparison.
<div><p>This study proposes the S-YOLOv10-ASI algorithm to improve the accuracy of tea identification and harvesting by robots, integrating a slice-assisted super-reasoning technique. The algorithm improves the partial structure of the YOLOv10 network through space-to-depth convolution....
Saved in:
| Main Author: | |
|---|---|
| Other Authors: | , , , , , , , |
| Published: |
2025
|
| Subjects: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1852018789777932288 |
|---|---|
| author | Chunhua Yang (346871) |
| author2 | Wenxia Yuan (3778042) Qiang Zhao (105948) Zejun Wang (4045592) Bowu Song (21647663) Xianqiu Dong (21647666) Yuandong Xiao (21647669) Shihao Zhang (5165825) Baijuan Wang (12588724) |
| author2_role | author author author author author author author author |
| author_facet | Chunhua Yang (346871) Wenxia Yuan (3778042) Qiang Zhao (105948) Zejun Wang (4045592) Bowu Song (21647663) Xianqiu Dong (21647666) Yuandong Xiao (21647669) Shihao Zhang (5165825) Baijuan Wang (12588724) |
| author_role | author |
| dc.creator.none.fl_str_mv | Chunhua Yang (346871) Wenxia Yuan (3778042) Qiang Zhao (105948) Zejun Wang (4045592) Bowu Song (21647663) Xianqiu Dong (21647666) Yuandong Xiao (21647669) Shihao Zhang (5165825) Baijuan Wang (12588724) |
| dc.date.none.fl_str_mv | 2025-07-02T17:45:09Z |
| dc.identifier.none.fl_str_mv | 10.1371/journal.pone.0325527.g011 |
| dc.relation.none.fl_str_mv | https://figshare.com/articles/figure/Different_model_detection_results_comparison_/29462681 |
| dc.rights.none.fl_str_mv | CC BY 4.0 info:eu-repo/semantics/openAccess |
| dc.subject.none.fl_str_mv | Genetics Biotechnology Developmental Biology Cancer Plant Biology Space Science Biological Sciences not elsewhere classified Chemical Sciences not elsewhere classified experimental results demonstrate approximately 10 %. loss function calculation anji white tea 10 %, 7 1 %, 6 classification loss 99 %, 69 %, target tea xlink "> validation set two leaves training set study proposes stage transmission single bud seen improvements resolves conflicts recognition ability reasoning technique reasoning algorithm partial structure one leaf one bud low resolution key layers fresh leaves depth convolution assisted super asi algorithm ap values adjacent layers |
| dc.title.none.fl_str_mv | Different model detection results comparison. |
| dc.type.none.fl_str_mv | Image Figure info:eu-repo/semantics/publishedVersion image |
| description | <div><p>This study proposes the S-YOLOv10-ASI algorithm to improve the accuracy of tea identification and harvesting by robots, integrating a slice-assisted super-reasoning technique. The algorithm improves the partial structure of the YOLOv10 network through space-to-depth convolution. The Progressive Feature Pyramid Network minimizes information loss during multi-stage transmission, enhances the saliency of key layers, resolves conflicts between objects, and improves the fusion of non-adjacent layers. Intersection over Union (IoU) is used to optimize the loss function calculation. The slice-assisted super-reasoning algorithm is integrated to improve the recognition ability of YOLOv10 network for long-distance and small-target tea. The experimental results demonstrate that when compared to YOLOv10, S-YOLOv10-ASI shows significant improvements across various metrics. Specifically, Bounding Box Regression Loss decreases by over 30% in the training set, while Classification Loss and Bounding Box Regression Loss drop by more than 60% in the validation set. Additionally, Distribution Focal Loss reduces by approximately 10%. Furthermore, Precision, Recall, and mAP have all increased by 7.1%, 6.69%, and 6.78% respectively. Moreover, the AP values for single bud, one bud and one leaf, and one bud and two leaves have seen improvements of 6.10%, 7.99%, and 8.28% respectively. The improved model effectively addresses challenges such as long-distance detection, small targets, and low resolution. It also offers high precision and recall, laying the foundation for the development of an Anji White Tea picking robot.</p></div> |
| eu_rights_str_mv | openAccess |
| id | Manara_da80fffa3ec75b55418dc2d86466bcfa |
| identifier_str_mv | 10.1371/journal.pone.0325527.g011 |
| network_acronym_str | Manara |
| network_name_str | ManaraRepo |
| oai_identifier_str | oai:figshare.com:article/29462681 |
| publishDate | 2025 |
| repository.mail.fl_str_mv | |
| repository.name.fl_str_mv | |
| repository_id_str | |
| rights_invalid_str_mv | CC BY 4.0 |
| spelling | Different model detection results comparison.Chunhua Yang (346871)Wenxia Yuan (3778042)Qiang Zhao (105948)Zejun Wang (4045592)Bowu Song (21647663)Xianqiu Dong (21647666)Yuandong Xiao (21647669)Shihao Zhang (5165825)Baijuan Wang (12588724)GeneticsBiotechnologyDevelopmental BiologyCancerPlant BiologySpace ScienceBiological Sciences not elsewhere classifiedChemical Sciences not elsewhere classifiedexperimental results demonstrateapproximately 10 %.loss function calculationanji white tea10 %, 71 %, 6classification loss99 %,69 %,target teaxlink ">validation settwo leavestraining setstudy proposesstage transmissionsingle budseen improvementsresolves conflictsrecognition abilityreasoning techniquereasoning algorithmpartial structureone leafone budlow resolutionkey layersfresh leavesdepth convolutionassisted superasi algorithmap valuesadjacent layers<div><p>This study proposes the S-YOLOv10-ASI algorithm to improve the accuracy of tea identification and harvesting by robots, integrating a slice-assisted super-reasoning technique. The algorithm improves the partial structure of the YOLOv10 network through space-to-depth convolution. The Progressive Feature Pyramid Network minimizes information loss during multi-stage transmission, enhances the saliency of key layers, resolves conflicts between objects, and improves the fusion of non-adjacent layers. Intersection over Union (IoU) is used to optimize the loss function calculation. The slice-assisted super-reasoning algorithm is integrated to improve the recognition ability of YOLOv10 network for long-distance and small-target tea. The experimental results demonstrate that when compared to YOLOv10, S-YOLOv10-ASI shows significant improvements across various metrics. Specifically, Bounding Box Regression Loss decreases by over 30% in the training set, while Classification Loss and Bounding Box Regression Loss drop by more than 60% in the validation set. Additionally, Distribution Focal Loss reduces by approximately 10%. Furthermore, Precision, Recall, and mAP have all increased by 7.1%, 6.69%, and 6.78% respectively. Moreover, the AP values for single bud, one bud and one leaf, and one bud and two leaves have seen improvements of 6.10%, 7.99%, and 8.28% respectively. The improved model effectively addresses challenges such as long-distance detection, small targets, and low resolution. It also offers high precision and recall, laying the foundation for the development of an Anji White Tea picking robot.</p></div>2025-07-02T17:45:09ZImageFigureinfo:eu-repo/semantics/publishedVersionimage10.1371/journal.pone.0325527.g011https://figshare.com/articles/figure/Different_model_detection_results_comparison_/29462681CC BY 4.0info:eu-repo/semantics/openAccessoai:figshare.com:article/294626812025-07-02T17:45:09Z |
| spellingShingle | Different model detection results comparison. Chunhua Yang (346871) Genetics Biotechnology Developmental Biology Cancer Plant Biology Space Science Biological Sciences not elsewhere classified Chemical Sciences not elsewhere classified experimental results demonstrate approximately 10 %. loss function calculation anji white tea 10 %, 7 1 %, 6 classification loss 99 %, 69 %, target tea xlink "> validation set two leaves training set study proposes stage transmission single bud seen improvements resolves conflicts recognition ability reasoning technique reasoning algorithm partial structure one leaf one bud low resolution key layers fresh leaves depth convolution assisted super asi algorithm ap values adjacent layers |
| status_str | publishedVersion |
| title | Different model detection results comparison. |
| title_full | Different model detection results comparison. |
| title_fullStr | Different model detection results comparison. |
| title_full_unstemmed | Different model detection results comparison. |
| title_short | Different model detection results comparison. |
| title_sort | Different model detection results comparison. |
| topic | Genetics Biotechnology Developmental Biology Cancer Plant Biology Space Science Biological Sciences not elsewhere classified Chemical Sciences not elsewhere classified experimental results demonstrate approximately 10 %. loss function calculation anji white tea 10 %, 7 1 %, 6 classification loss 99 %, 69 %, target tea xlink "> validation set two leaves training set study proposes stage transmission single bud seen improvements resolves conflicts recognition ability reasoning technique reasoning algorithm partial structure one leaf one bud low resolution key layers fresh leaves depth convolution assisted super asi algorithm ap values adjacent layers |