Skip to content
DraconiticResearch
Search Sources API
226,742 studies31 sourcesUpdated 40 min ago
Peer-reviewedCC BY 4.0Frontiers in Plant Science

Real-time mulberry fruit ripeness recognition for selective harvesting using edge-deployable object detection: a single-orchard case study

Abstract

IntroductionMulberry is a seasonal cash crop with a short ripening window, where delayed harvesting readily leads to fruit deterioration and economic losses. Real-time ripeness recognition is therefore important for selective harvesting. Existing detection models face challenges including excessive model parameters, low discriminability between immature fruits and background foliage, and severe occlusion caused by clustered fruits.MethodsWe present Mulberry-Net, a lightweight three-class mulberry ripeness detector that adapts and systematically evaluates three established components: Partial Convolution integrated into the C2f bottleneck to form the Cross-Stage Feature Compression and Cooperation (CSFCC) module; Mixed Local Channel Attention (MLCA) placed after multi-scale feature concatenation in the neck; and an occlusion-aware training strategy based on Focal-DIoU loss. The components target computational cost, fruit-background discrimination, and localisation under occlusion, respectively.ResultsMulberry-Net achieves an [email protected] of 97.80% with 2.41M parameters, improving on the YOLOv8n baseline by 5.10 percentage points while reducing parameters by 19.9%. The largest class-specific gain occurs for green fruits (+8.60 percentage points). Five-fold tree-level cross-validation yields 96.02 +/- 0.74% [email protected]. Three-class ripeness accuracy on correctly localised fruits reaches 96.83%, with residual grading errors concentrated at the color-turning/mature boundary and missed detections concentrated in the green stage. A controlled comparison against four alternative lightweight convolution designs at the same bottleneck position shows that the adopted configuration attains the highest accuracy at comparable computational cost.DiscussionThe results support the feasibility of lightweight three-class mulberry ripeness detection for selective harvesting under the conditions studied. Because all data originate from a single orchard, cultivar, and acquisition device, the reported figures should be interpreted as single-site feasibility evidence rather than established generalisation across cultivars and growing regions.

Related

  1. Edge-deployable greenhouse tomato cluster harvesting robot integrating YOLOv8n-BiFPN-WIoU and ROS-based autonomous control
    Fanzhao Meng, Zhen Wang, Huan Wang +7
    Frontiers in plant sciencePeer-reviewedCC BYFull textPlant biologyEurope PMCCrossref
  2. A novel immunochromatographic strip for rapid and field-deployable detection of pigeon adenovirus type 1
    Jingyi Wang, Yaning Gao, Yupeng Fang +7
    Poultry SciencePeer-reviewedCC BY-NC-ND 4.0Full textZoology & animal behaviourPMC
  3. A lightweight edge-deployable detection method for sugarcane stem nodes in intelligent seed cane cutting equipment
    Zhaojun Niu, Huadong Xiao, Zebin Wu +8
    Frontiers in plant sciencePeer-reviewedCC BYFull textPlant biologyEurope PMCCrossref
  4. Multi-scale lodging in rice and wheat: bridging biomechanical traits, phenotypic detection, and intelligent harvesting systems
    Tiantian Jing, Xianfei Xia, Zhong Tang +2
    Frontiers in plant sciencePeer-reviewedCC BYFull textPlant biologyEurope PMC
  5. Field-deployable seed health diagnostic assay: RT-RPA-based detection of turnip mosaic virus in mustard leaf and seed lots
    Damini Diksha, Pankhuri Singhal, Aakansha Manav +6
    BMC Plant BiologyPeer-reviewedCC BY-NC-ND 4.0Full textPlant biologyPMC
BibTeX
@article{chan2026realtime,
  title = {{Real-time mulberry fruit ripeness recognition for selective harvesting using edge-deployable object detection: a single-orchard case study}},
  author = {Zhongbin Chan and Xin Zhao and Xinwei Zhang and Bojian Yu},
  journal = {Frontiers in Plant Science},
  year = {2026},
  month = nov,
  doi = {10.3389/fpls.2026.1942552},
  url = {https://www.frontiersin.org/articles/10.3389/fpls.2026.1942552},
  note = {CC BY 4.0},
}
RIS
TY  - JOUR
TI  - Real-time mulberry fruit ripeness recognition for selective harvesting using edge-deployable object detection: a single-orchard case study
AU  - Zhongbin Chan
AU  - Xin Zhao
AU  - Xinwei Zhang
AU  - Bojian Yu
PY  - 2026
DA  - 2026/11/16
JO  - Frontiers in Plant Science
DO  - 10.3389/fpls.2026.1942552
UR  - https://www.frontiersin.org/articles/10.3389/fpls.2026.1942552
AB  - IntroductionMulberry is a seasonal cash crop with a short ripening window, where delayed harvesting readily leads to fruit deterioration and economic losses. Real-time ripeness recognition is therefore important for selective harvesting. Existing detection models face challenges including excessive model parameters, low discriminability between immature fruits and background foliage, and severe occlusion caused by clustered fruits.MethodsWe present Mulberry-Net, a lightweight three-class mulberry ripeness detector that adapts and systematically evaluates three established components: Partial Convolution integrated into the C2f bottleneck to form the Cross-Stage Feature Compression and Cooperation (CSFCC) module; Mixed Local Channel Attention (MLCA) placed after multi-scale feature concatenation in the neck; and an occlusion-aware training strategy based on Focal-DIoU loss. The components target computational cost, fruit-background discrimination, and localisation under occlusion, respectively.ResultsMulberry-Net achieves an [email protected] of 97.80% with 2.41M parameters, improving on the YOLOv8n baseline by 5.10 percentage points while reducing parameters by 19.9%. The largest class-specific gain occurs for green fruits (+8.60 percentage points). Five-fold tree-level cross-validation yields 96.02 +/- 0.74% [email protected]. Three-class ripeness accuracy on correctly localised fruits reaches 96.83%, with residual grading errors concentrated at the color-turning/mature boundary and missed detections concentrated in the green stage. A controlled comparison against four alternative lightweight convolution designs at the same bottleneck position shows that the adopted configuration attains the highest accuracy at comparable computational cost.DiscussionThe results support the feasibility of lightweight three-class mulberry ripeness detection for selective harvesting under the conditions studied. Because all data originate from a single orchard, cultivar, and acquisition device, the reported figures should be interpreted as single-site feasibility evidence rather than established generalisation across cultivars and growing regions.
SN  - 1664-462X
ER  -