Green Fruit‐Stem Pairing and Clustering for Machine Vision System in Robotic Thinning of Apples DOI Creative Commons

Magni Hussain,

Long He,

James R. Schupp

и другие.

Journal of Field Robotics, Год журнала: 2024, Номер unknown

Опубликована: Ноя. 5, 2024

ABSTRACT Apples are one of the most highly‐valued specialty crops in United States. Recent labor shortages have made crop production difficult for fruit growers, including task green thinning. Current methods hand, chemical, and mechanical thinning impose tradeoffs between selectivity, cost, tree damage, speed. A robotic system could potentially selectively thin a quick, cost‐effective, non‐damaging manner. The machine vision would be critical component thinning, not only need to perform detection/segmentation, but also fruit‐stem pairing clustering facilitate proper decision‐making neural network‐based stem algorithm was devised evaluated; an LSTM‐based compared density‐based algorithm, OPTICS. algorithms were evaluated on image data set consisting GoldRush, Fuji, Golden Delicious cultivars at stage, with evaluations overall performance, cultivar‐wise cluster size‐specific feature importance. For pairing, achieved AP 81.4% all fruits stems, that reached 90.6% when stems labeled angles considered. clustering, success rate 68.4%, whereas OPTICS obtained 50.9%. will further implemented pipeline future system, as well integrate use point clouds other 3D orchard information improve performance.

Язык: Английский

YOLO-Peach: A High-Performance Lightweight YOLOv8s-Based Model for Accurate Recognition and Enumeration of Peach Seedling Fruits DOI Creative Commons
Yi Shi,

Shunhao Qing,

Long Zhao

и другие.

Agronomy, Год журнала: 2024, Номер 14(8), С. 1628 - 1628

Опубликована: Июль 25, 2024

The identification and enumeration of peach seedling fruits are pivotal in the realm precision agriculture, greatly influencing both yield estimation agronomic practices. This study introduces an innovative, lightweight YOLOv8 model for automatic detection quantification fruits, designated as YOLO-Peach, to bolster scientific rigor operational efficiency orchard management. Traditional methods, which labor-intensive error-prone, have been superseded by this advancement. A comprehensive dataset was meticulously curated, capturing rich characteristics diversity through high-resolution imagery at various times locations, followed meticulous preprocessing ensure data quality. YOLOv8s underwent a series optimizations, including integration MobileNetV3 its backbone, p2BiFPN architecture, spatial channel reconstruction convolution, coordinate attention mechanism, all significantly bolstered model’s capability detect small targets with precision. YOLO-Peach excels accuracy, evidenced recall 0.979, along mAP50 0.993 mAP50-95 0.867, indicating superior sapling efficient computational performance. findings underscore efficacy practicality context fruit recognition. Ablation studies shed light on indispensable role each component, streamlining complexity load, while ScConv convolutions, mechanism collectively enhanced feature extraction minute targets. implications research profound, offering novel approach recognition serving blueprint young other species. work holds significant theoretical practical value, propelling forward broader field agricultural automation.

Язык: Английский

Процитировано

6

A Novel Multinozzle Targeting Pollination Robot for Clustered Kiwifruit Flowers Based on Air–Liquid Dual‐Flow Spraying DOI Open Access

Changqing Gao,

Leilei He,

Yezhang Ding

и другие.

Journal of Field Robotics, Год журнала: 2025, Номер unknown

Опубликована: Янв. 19, 2025

ABSTRACT Manual pollination of kiwifruit flowers is a labor‐intensive work that highly desired to be replaced by robotic operations. In this research, robot was developed achieve precision clustered in the orchard. The consists five systems, including multinozzle end‐effector, mechanical arm, vision system, crawler‐type chassis, and control system. can select preferential then target their pistil pollination. First, statistical analysis dimensions flower clusters individual conducted fit normal distribution curves, which guided design spray coverage combination intervals for end‐effector. Second, optimal parameters were determined based on three‐factor, five‐level quadratic orthogonal experiment, is, air pressure 70.4 kPa, rate flow 86.0 mL/min, distance 27.8 cm. A targeted strategy selection structure Field experiments commercial orchard evaluate its feasibility performance, an average success targeting 93.4% at speed 1.0 s per achieved. Furthermore, compared with artificial assisted methods, it improve utilization pollen consumption 0.20 g every 60 fruit set 88.9%. validations demonstrated efficiently pollinate save pollen.

Язык: Английский

Процитировано

0

A vision-based robotic system for precision pollination of apples DOI Creative Commons
Uddhav Bhattarai, Ranjan Sapkota,

Safal Kshetri

и другие.

Computers and Electronics in Agriculture, Год журнала: 2025, Номер 234, С. 110158 - 110158

Опубликована: Март 9, 2025

Язык: Английский

Процитировано

0

Robotic precision thinning in apple production – using optimization and Bayesian modeling to assess potentials of automation in horticulture DOI
Johannes Kopton,

L. Zimmermann,

Eike Luedeling

и другие.

Acta Horticulturae, Год журнала: 2025, Номер 1425, С. 239 - 246

Опубликована: Март 1, 2025

Язык: Английский

Процитировано

0

Applications of the Internet of Things on Agriculture: Review and Future Apple Fruit Comprehensive Automation System DOI
Mustafa Mhamed, Zhao Zhang

Smart agriculture, Год журнала: 2025, Номер unknown, С. 45 - 68

Опубликована: Янв. 1, 2025

Язык: Английский

Процитировано

0

AI Technologies for Apple Leaf Diseases Identification: Scientometric Analysis DOI
Mustafa Mhamed, Zhao Zhang

Smart agriculture, Год журнала: 2025, Номер unknown, С. 137 - 161

Опубликована: Янв. 1, 2025

Язык: Английский

Процитировано

0

Research Progress on Thinning Equipment in Orchards: A Review DOI
Xin Jiao,

Congjue Tan,

Mustafa Mhamed

и другие.

Smart agriculture, Год журнала: 2025, Номер unknown, С. 23 - 44

Опубликована: Янв. 1, 2025

Язык: Английский

Процитировано

0

Comparison of robotic precision thinning system and commercial air-blast sprayer for flower thinning on apple trees DOI

C. Andergassen,

Emiliano Bruni,

D. Pichler

и другие.

Acta Horticulturae, Год журнала: 2024, Номер 1395, С. 369 - 372

Опубликована: Май 1, 2024

Язык: Английский

Процитировано

0

Robotics for tree fruit orchards DOI
Manoj Karkee

Acta Horticulturae, Год журнала: 2024, Номер 1395, С. 359 - 368

Опубликована: Май 1, 2024

Язык: Английский

Процитировано

0

Green Fruit‐Stem Pairing and Clustering for Machine Vision System in Robotic Thinning of Apples DOI Creative Commons

Magni Hussain,

Long He,

James R. Schupp

и другие.

Journal of Field Robotics, Год журнала: 2024, Номер unknown

Опубликована: Ноя. 5, 2024

ABSTRACT Apples are one of the most highly‐valued specialty crops in United States. Recent labor shortages have made crop production difficult for fruit growers, including task green thinning. Current methods hand, chemical, and mechanical thinning impose tradeoffs between selectivity, cost, tree damage, speed. A robotic system could potentially selectively thin a quick, cost‐effective, non‐damaging manner. The machine vision would be critical component thinning, not only need to perform detection/segmentation, but also fruit‐stem pairing clustering facilitate proper decision‐making neural network‐based stem algorithm was devised evaluated; an LSTM‐based compared density‐based algorithm, OPTICS. algorithms were evaluated on image data set consisting GoldRush, Fuji, Golden Delicious cultivars at stage, with evaluations overall performance, cultivar‐wise cluster size‐specific feature importance. For pairing, achieved AP 81.4% all fruits stems, that reached 90.6% when stems labeled angles considered. clustering, success rate 68.4%, whereas OPTICS obtained 50.9%. will further implemented pipeline future system, as well integrate use point clouds other 3D orchard information improve performance.

Язык: Английский

Процитировано

0