Lightweight Detection Method for Young Grape Cluster Fruits Based on SAW YOLO v8n

被引：0
|
作者：
Zhang, Chuandong [1 ]
Gao, Peng [2 ]
Qi, Lu [1 ]
Ding, Huali [1 ]
机构：
[1] School qf Computer Science and Engineering, JVning University, Qufu
[2] School of Cyber Science and Engineering, Qufu Normal University, Qufu
来源：
Nongye Jixie Xuebao/Transactions of the Chinese Society for Agricultural Machinery | 2024年 / 55卷 / 10期
关键词：
fruit thinning; object detection; shuffle attention; Wise — IoU Löss; YOLO v8n; young grape Cluster fruits;
D O I：
10.6041/j.issn.1000-1298.2024.10.027
中图分类号：
学科分类号：
摘要：
The detection of young grape eluster fruits is challenging due to the influenae of background color, oeelusion, and lighting variations. To achieve robust detection of young grape Cluster fruits for the varying conditions, an improved YOLO v8n model that integrated shuffle attention (SA) mechanism was proposed in the work. By incorporating SA mechanism into the Neck network of the YOLO v8n model, the multi-scale feature fusion ability of the network was enhanced, the feature Information representation of the detection target was improved, and other irrelevant information was suppressed, improving the accuracy of the detection network, which achieved efficient and accurate detection of young grape eluster fruits without significantly increasing network depth and memory overhead. Wise intersection over union loss (Wise - IoU Löss) with the dynamic nonmonotonic focusing mechanism was taken as the bounding box regression loss funetion, to accelerate the network convergence for the better detection accuracy of the model. Herein, a Grape dataset was construeted, which comprised 3 780 images of young grape eluster fruits in complex scenarios along with corresponding annotation files. Training and testing results of the SAW- YOLO v8n model on this dataset showed that the precision (P), recall (R), mean average precision (mAP), and Fl score of the young grape Cluster fruit detection algorithm based on SAW _ YOLO v8n were 92. 80%, 91. 30%, 96. 10%, and 92. 04%, respectively, where the detection speed was 140. 85 f/s, and the model size was only 6. 20 MB. Compared with that of SSD, YOLO v5s, YOLO v6n, YOLO v7-tiny, and YOLO v8n, the mAP was increased by 16.06%, 1.05%, 1.48%, 0.84% and 0. 73%, respectively, and Fl scores were increased by 24. 85%, 1. 43%, 1. 43%, 1. 09% and 1. 60%, respectively, and the model weights were reduced by 93. 16%, 56. 94%, 37. 63%, 47. 00%, and 0, respectively, which was the smallest among all models and had obvious advantages in lightweight and high accuracy. Moreover, the young grape Cluster fruits detection with different degrees of occlusion and lighting conditions were also explored, and the result showed that the young grape Cluster fruit detection method based on SAW — YOLO v8n can adapt to different occlusion and lighting changes, and had good robustness. In summary, SAW — YOLO v8n not only met the requirements of high-precision, high-speed, and lightweight detection of young grape Cluster fruits, but also had strong robustness and real-time Performance. © 2024 Chinese Society of Agricultural Machinery. All rights reserved.
引用
收藏
页码：286 / 294
页数：8
相关论文
共 28 条
  • [1] LIU S, LI X, WU H, Et al., A robust automated flower estimation System for grape vines, Biosystems Engineering, 172, pp. 110-123, (2018)
  • [2] DU W, ZHU Y, LI S, Et al., Spikelets deteetion of table grape before thinning based on improved YOLO V5s and Kmeans under the complex environment[J], Computers and Electronics in Agrieulture, 203, (2022)
  • [3] RUDOLPH R, HERZOG K, TOPFER R, Et al., Efficient identification, loealization and quantification of grapevine inflorescences in unprepared field images using fully eonvolutional networks, Vitis Journal of Grapevine Research, 58, 3, pp. 95-104, (2019)
  • [4] PALACIOS F, BUENO G, SALIDO J, Et al., Automated grapevine flower deteetion and quantification method based on Computer vision and deep learning from on-the-go imaging using a mobile sensing platform under field conditions, Computers and Electronics in Agrieulture, 178, (2020)
  • [5] LIU Ping, ZHU Yanjun, ZHANG Tongxun, Et al., Algorithm for recognition and image segmentation of overlapping grape Cluster in natural environment[J], Transactions of the CSAE, 36, 6, pp. 161-169, (2020)
  • [6] SONG Huaibo, SHANG Yuying, HE Dongjian, Review on deep learning teehnology for fruit target recognition [J], Transactions of the Chinese Society for Agricultural Machinery, 54, 1, pp. 1-19, (2023)
  • [7] SANTOS T T, SOUZA L L D, SANTOS A A D, Et al., Grape deteetion, segmentation, and tracking using deep neural networks and three-dimensional association[J], Computers and Electronics in Agrieulture, 170, (2020)
  • [8] SONG Huaibo, JIANG Mei, WANG Yunfei, Et al., Efficient detection method for young apples based on the fusion of convolutional neural network and visual attention meehanism [J], Transactions of the GSAE, 37, 9, pp. 297-303, (2021)
  • [9] SONG Huaibo, MA Baoling, SHANG Yuying, Et al., Detection of young apple fruits based on YOLO v7 - ECA model [J], Transactions of the Chinese Society for Agrieultural Maehinery, 54, 6, pp. 233-242, (2023)
  • [10] SONG Huaibo, WANG Ya'nan, WANG Yunfei, Et al., Camellia oleifera fruit detection in natural scene based on YOLO v5s[J], Transactions of the Chinese Society for Agrieultural Maehinery, 53, 7, pp. 234-242, (2022)
123