EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Accurate and fast implementation of soybean pod counting and localization from high-resolution image

Zhenghong Yu, Yangxu Wang, Jianxiong Ye, Shengjie Liufu, Dunlu Lu, Xiuli Zhu, Zhongming Yang, Qingji Tan

Frontiers in Plant Science · 20.02.2024 · 15

DOI: 10.3389/fpls.2024.1320109 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Frontiers in Plant Science, online erschienen am 20.02.2024. Verfasst von Zhenghong Yu und 7 weiteren Autor:innen. Beteiligte Einrichtung: College of Robotics, Guangdong Polytechnic of Science and Technology, Zhuhai, China. Die Arbeit wurde bisher 20-mal zitiert (Crossref) und nennt 38 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Introduction

Soybean pod count is one of the crucial indicators of soybean yield. Nevertheless, due to the challenges associated with counting pods, such as crowded and uneven pod distribution, existing pod counting models prioritize accuracy over efficiency, which does not meet the requirements for lightweight and real-time tasks.

Methods

To address this goal, we have designed a deep convolutional network called PodNet. It employs a lightweight encoder and an efficient decoder that effectively decodes both shallow and deep information, alleviating the indirect interactions caused by information loss and degradation between non-adjacent levels.

Results

We utilized a high-resolution dataset of soybean pods from field harvesting to evaluate the model’s generalization ability. Through experimental comparisons between manual counting and model yield estimation, we confirmed the effectiveness of the PodNet model. The experimental results indicate that PodNet achieves an R 2 of 0.95 for the prediction of soybean pod quantities compared to ground truth, with only 2.48M parameters, which is an order of magnitude lower than the current SOTA model YOLO POD, and the FPS is much higher than YOLO POD.

Discussion

Compared to advanced computer vision methods, PodNet significantly enhances efficiency with almost no sacrifice in accuracy. Its lightweight architecture and high FPS make it suitable for real-time applications, providing a new solution for counting and locating dense objects.

Schlagwörter

Stichwörter der Autor:innen

computer visionconvolutional networkcounting and locatingdense objectssoybean pod

Autor:innen und Institutionen

  • College of Robotics, Guangdong Polytechnic of Science and Technology, Zhuhai, China.
  • Department of Network Technology, Guangzhou Institute of Software Engineering, Conghua, China.
  • School of Electronics and Information Engineering, Wuyi University, Jiangmen, China.
  • College of Business, Guangzhou College of Technology and Business, Foshan, China.
  • School of Mechanical Engineering, Guangxi University, Nanning, China.

Zitieren

APA

Yu, Z., Wang, Y., Ye, J., Liufu, S., Lu, D., Zhu, X., Yang, Z., & Tan, Q. (2024). Accurate and fast implementation of soybean pod counting and localization from high-resolution image. Frontiers in Plant Science, 15. https://doi.org/10.3389/fpls.2024.1320109

Vancouver

Yu Z, Wang Y, Ye J, Liufu S, Lu D, Zhu X, et al. Accurate and fast implementation of soybean pod counting and localization from high-resolution image. Frontiers in Plant Science. 2024;15. doi:10.3389/fpls.2024.1320109

BibTeX
@article{Yu2024,
  author = {Yu, Zhenghong and Wang, Yangxu and Ye, Jianxiong and Liufu, Shengjie and Lu, Dunlu and Zhu, Xiuli and Yang, Zhongming and Tan, Qingji},
  title = {Accurate and fast implementation of soybean pod counting and localization from high-resolution image},
  journal = {Frontiers in Plant Science},
  year = {2024},
  volume = {15},
  doi = {10.3389/fpls.2024.1320109}
}
Kostenlos starten →