A Method of Grasping Detection for Kiwifruit Harvesting Robot Based on Deep Learning

Ma, Li; He, Zhi; Zhu, Yuntian; Jia, Liangsheng; Wang, Yinchu; Ding, Xinting; Cui, Yongjie

doi:10.3390/agronomy12123096

Cited by 9 publications

(2 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Additionally, a maximum of two adjacent fruits can be found per fruit. The fruits within aggregation clusters display irregular distribution and usually comprise 4-6 fruits (Ma et al, 2022). The spatial distribution of kiwifruit fruit in the picking area affects the efficiency of the end-effector picking operation (Li et al, 2022).…”

Section: Kiwifruit Distribution Characteristicsmentioning

confidence: 99%

Design and Experiment of Multi-Fruit Gripping

FU,

CAI,

GUO

et al. 2024

INMATEH

View full text Add to dashboard Cite

Kiwifruit picking robots can replace manual labor for mechanized kiwifruit harvesting. However, existing picking robots encounter issues such as low separation efficiency of fruit stalks, poor stability of fruit gripping, and inaccurate identification when picking kiwifruit clusters. In response, a multi-fruit picking end-effector was designed to pick clusters of kiwifruit efficiently based on their distribution characteristics. The gripping range of the gripping device was determined based on the parameters of the spatial distribution of the fruit clusters. A multi-fruit gripping mechanics model was constructed, and the gripping force was analyzed to ensure efficient and stable fruit picking. Critical parameters of the fruit stalk separation device were determined through kinematic trajectory analysis to improve the separation efficiency of fruit stalks. Additionally, a dual-sensor fusion recognition method was proposed to identify fruit cluster locations accurately. The results of the picking experiment demonstrate that the end-effector can pick fruits in an average time of 8.28 s per cluster, with a net fruit-picking rate of 87.5% and a fruit damage rate of 7.5%. The end-effector shows a positive picking effect on kiwifruit fruits distributed in clusters. This study can serve as a reference for the development of kiwifruit-picking robots.

show abstract

Section: Kiwifruit Distribution Characteristicsmentioning

confidence: 99%

Design and Experiment of Multi-Fruit Gripping

FU,

CAI,

GUO

et al. 2024

INMATEH

View full text Add to dashboard Cite

show abstract

“…Accurate and rapid detection of tea canopy shoots in complex field environments is one of the crucial technologies for intelligent picking platforms. Computer vision technology has been widely applied in target detection of various fruits and vegetables, such as apple [5], tomato [6], strawberry [7], kiwifruit [8], and grape [9]. The primary techniques used for tea shoot detection involve traditional image processing and deep learning methods.…”

Section: Introductionmentioning

confidence: 99%

TS-YOLO: An All-Day and Lightweight Tea Canopy Shoots Detection Model

Zhang

Zhao

et al. 2023

Agronomy

View full text Add to dashboard Cite

Accurate and rapid detection of tea shoots within the tea canopy is essential for achieving the automatic picking of famous tea. The current detection models suffer from two main issues: low inference speed and difficulty in deployment on movable platforms, which constrain the development of intelligent tea picking equipment. Furthermore, the detection of tea canopy shoots is currently limited to natural daylight conditions, with no reported studies on detecting tea shoots under artificial light during the nighttime. Developing an all-day tea picking platform would significantly improve the efficiency of tea picking. In view of these problems, the research objective was to propose an all-day lightweight detection model for tea canopy shoots (TS-YOLO) based on YOLOv4. Firstly, image datasets of tea canopy shoots sample were collected under low light (6:30–7:30 and 18:30–19:30), medium light (8:00–9:00 and 17:00–18:00), high light (11:00–15:00), and artificial light at night. Then, the feature extraction network of YOLOv4 and the standard convolution of the entire network were replaced with the lightweight neural network MobilenetV3 and the depth-wise separable convolution. Finally, to compensate for the lack of feature extraction ability in the lightweight neural network, a deformable convolutional layer and coordinate attention modules were added to the network. The results showed that the improved model size was 11.78 M, 18.30% of that of YOLOv4, and the detection speed was improved by 11.68 FPS. The detection accuracy, recall, and AP of tea canopy shoots under different light conditions were 85.35%, 78.42%, and 82.12%, respectively, which were 1.08%, 12.52%, and 8.20% higher than MobileNetV3-YOLOv4, respectively. The developed lightweight model could effectively and rapidly detect tea canopy shoots under all-day light conditions, which provides the potential to develop an all-day intelligent tea picking platform.

show abstract

Research on the visual location method for strawberry picking points under complex conditions based on composite models

Xie,

Zhang,

Zhang

et al. 2024

J Sci Food Agric

View full text Add to dashboard Cite

BackgroundStrawberry, being an important economic crop, requires a large amount of human labor for harvesting operations. Efficient and non‐destructive harvesting by strawberry harvesting robots requires the precise location of the picking points. Current algorithms for locating picking points encounter significant issues with location errors and minimal effective information in complex situations.ResultsTo improve the accuracy of the location of picking points, this study proposes a visual location method based on composite models. This method employs object detection and instance segmentation models to detect fruits and segment peduncles sequentially, thereby enabling the identification of picking points and inclination on the peduncle. Different object detection algorithms and instance segmentation models were validated to explore the optimal model combination, and the Convolutional Block Attention Module (CBAM) was integrated into YOLOv8s‐seg to construct YOLOv8s‐seg‐CBAM. Test results show that the composite model built with YOLOv8s and YOLOv8s‐seg‐CBAM achieved a peduncle detection accuracy of 86.2%, with an inference time of 30.6 ms per image.ConclusionThe picking point visual location method based on YOLOv8s and YOLOv8s‐seg‐CBAM composite models can better balance accuracy and efficiency and can provide more accurate guidance for automated harvesting. © 2024 Society of Chemical Industry.

show abstract

A Method of Grasping Detection for Kiwifruit Harvesting Robot Based on Deep Learning

Cited by 9 publications

References 22 publications

Design and Experiment of Multi-Fruit Gripping

Design and Experiment of Multi-Fruit Gripping

TS-YOLO: An All-Day and Lightweight Tea Canopy Shoots Detection Model

Research on the visual location method for strawberry picking points under complex conditions based on composite models

Contact Info

Product

Resources

About