D<sup>3</sup>-YOLOv10: Improved YOLOv10-Based Lightweight Tomato Detection Algorithm Under Facility Scenario
Accurate and efficient tomato detection is one of the key techniques for intelligent automatic picking in the area of precision agriculture. However, under the facility scenario, existing detection algorithms still have challenging problems such as weak feature extraction ability for occlusion condi...
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2024-12-01
|
| Series: | Agriculture |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2077-0472/14/12/2268 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1850050666949509120 |
|---|---|
| author | Ao Li Chunrui Wang Tongtong Ji Qiyang Wang Tianxue Zhang |
| author_facet | Ao Li Chunrui Wang Tongtong Ji Qiyang Wang Tianxue Zhang |
| author_sort | Ao Li |
| collection | DOAJ |
| description | Accurate and efficient tomato detection is one of the key techniques for intelligent automatic picking in the area of precision agriculture. However, under the facility scenario, existing detection algorithms still have challenging problems such as weak feature extraction ability for occlusion conditions and different fruit sizes, low accuracy on edge location, and heavy model parameters. To address these problems, this paper proposed D<sup>3</sup>-YOLOv10, a lightweight YOLOv10-based detection framework. Initially, a compact dynamic faster network (DyFasterNet) was developed, where multiple adaptive convolution kernels are aggregated to extract local effective features for fruit size adaption. Additionally, the deformable large kernel attention mechanism (D-LKA) was designed for the terminal phase of the neck network by adaptively adjusting the receptive field to focus on irregular tomato deformations and occlusions. Then, to further improve detection boundary accuracy and convergence, a dynamic FM-WIoU regression loss with a scaling factor was proposed. Finally, a knowledge distillation scheme using semantic frequency prompts was developed to optimize the model for lightweight deployment in practical applications. We evaluated the proposed framework using a self-made tomato dataset and designed a two-stage category balancing method based on diffusion models to address the sample class-imbalanced issue. The experimental results demonstrated that the D<sup>3</sup>-YOLOv10 model achieved an <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><mrow><mi>m</mi><mi>A</mi><msub><mi>P</mi><mrow><mn>0.5</mn></mrow></msub></mrow></semantics></math></inline-formula> of 91.8%, with a substantial reduction of 54.0% in parameters and 64.9% in FLOPs, compared to the benchmark model. Meanwhile, the detection speed of 80.1 FPS more effectively meets the demand for real-time tomato detection. This study can effectively contribute to the advancement of smart agriculture research on the detection of fruit targets. |
| format | Article |
| id | doaj-art-e0ae85131a1f401fb69b2bb3ad230a8d |
| institution | DOAJ |
| issn | 2077-0472 |
| language | English |
| publishDate | 2024-12-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Agriculture |
| spelling | doaj-art-e0ae85131a1f401fb69b2bb3ad230a8d2025-08-20T02:53:22ZengMDPI AGAgriculture2077-04722024-12-011412226810.3390/agriculture14122268D<sup>3</sup>-YOLOv10: Improved YOLOv10-Based Lightweight Tomato Detection Algorithm Under Facility ScenarioAo Li0Chunrui Wang1Tongtong Ji2Qiyang Wang3Tianxue Zhang4School of Computer Science and Technology, Harbin University of Science and Technology, Harbin 150080, ChinaSchool of Computer Science and Technology, Harbin University of Science and Technology, Harbin 150080, ChinaSchool of Computer Science and Technology, Harbin University of Science and Technology, Harbin 150080, ChinaSchool of Agricultural Engineering, Jiangsu University, Zhenjiang 212013, ChinaSchool of Mechanical Engineering and Automation, Beihang University, Beijing 100191, ChinaAccurate and efficient tomato detection is one of the key techniques for intelligent automatic picking in the area of precision agriculture. However, under the facility scenario, existing detection algorithms still have challenging problems such as weak feature extraction ability for occlusion conditions and different fruit sizes, low accuracy on edge location, and heavy model parameters. To address these problems, this paper proposed D<sup>3</sup>-YOLOv10, a lightweight YOLOv10-based detection framework. Initially, a compact dynamic faster network (DyFasterNet) was developed, where multiple adaptive convolution kernels are aggregated to extract local effective features for fruit size adaption. Additionally, the deformable large kernel attention mechanism (D-LKA) was designed for the terminal phase of the neck network by adaptively adjusting the receptive field to focus on irregular tomato deformations and occlusions. Then, to further improve detection boundary accuracy and convergence, a dynamic FM-WIoU regression loss with a scaling factor was proposed. Finally, a knowledge distillation scheme using semantic frequency prompts was developed to optimize the model for lightweight deployment in practical applications. We evaluated the proposed framework using a self-made tomato dataset and designed a two-stage category balancing method based on diffusion models to address the sample class-imbalanced issue. The experimental results demonstrated that the D<sup>3</sup>-YOLOv10 model achieved an <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><mrow><mi>m</mi><mi>A</mi><msub><mi>P</mi><mrow><mn>0.5</mn></mrow></msub></mrow></semantics></math></inline-formula> of 91.8%, with a substantial reduction of 54.0% in parameters and 64.9% in FLOPs, compared to the benchmark model. Meanwhile, the detection speed of 80.1 FPS more effectively meets the demand for real-time tomato detection. This study can effectively contribute to the advancement of smart agriculture research on the detection of fruit targets.https://www.mdpi.com/2077-0472/14/12/2268tomato detectionYOLOv10occlusion recognitionattention mechanismknowledge distillation |
| spellingShingle | Ao Li Chunrui Wang Tongtong Ji Qiyang Wang Tianxue Zhang D<sup>3</sup>-YOLOv10: Improved YOLOv10-Based Lightweight Tomato Detection Algorithm Under Facility Scenario Agriculture tomato detection YOLOv10 occlusion recognition attention mechanism knowledge distillation |
| title | D<sup>3</sup>-YOLOv10: Improved YOLOv10-Based Lightweight Tomato Detection Algorithm Under Facility Scenario |
| title_full | D<sup>3</sup>-YOLOv10: Improved YOLOv10-Based Lightweight Tomato Detection Algorithm Under Facility Scenario |
| title_fullStr | D<sup>3</sup>-YOLOv10: Improved YOLOv10-Based Lightweight Tomato Detection Algorithm Under Facility Scenario |
| title_full_unstemmed | D<sup>3</sup>-YOLOv10: Improved YOLOv10-Based Lightweight Tomato Detection Algorithm Under Facility Scenario |
| title_short | D<sup>3</sup>-YOLOv10: Improved YOLOv10-Based Lightweight Tomato Detection Algorithm Under Facility Scenario |
| title_sort | d sup 3 sup yolov10 improved yolov10 based lightweight tomato detection algorithm under facility scenario |
| topic | tomato detection YOLOv10 occlusion recognition attention mechanism knowledge distillation |
| url | https://www.mdpi.com/2077-0472/14/12/2268 |
| work_keys_str_mv | AT aoli dsup3supyolov10improvedyolov10basedlightweighttomatodetectionalgorithmunderfacilityscenario AT chunruiwang dsup3supyolov10improvedyolov10basedlightweighttomatodetectionalgorithmunderfacilityscenario AT tongtongji dsup3supyolov10improvedyolov10basedlightweighttomatodetectionalgorithmunderfacilityscenario AT qiyangwang dsup3supyolov10improvedyolov10basedlightweighttomatodetectionalgorithmunderfacilityscenario AT tianxuezhang dsup3supyolov10improvedyolov10basedlightweighttomatodetectionalgorithmunderfacilityscenario |