EUVIMEDEUVIMED entdecken
Open Access · CC BY-NC-ND 4.0

An efficient and lightweight banana detection and localization system based on deep CNNs for agricultural robots

Zhenhui Zheng, Ling Chen, Lijiao Wei, Weihua Huang, Dongjie Du, Guoquan Qin, Zhou Yang, Shuo Wang

Smart Agricultural Technology · 2024-12 · 9 · 100550

DOI: 10.1016/j.atech.2024.100550 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Smart Agricultural Technology, 2024-12. Verfasst von Zhenhui Zheng und 7 weiteren Autor:innen. Die Arbeit wurde bisher 5-mal zitiert (Crossref) und nennt 29 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC-ND 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Accurate detection and positioning of fruits in natural environments is a key step for fruit picking robots to achieve precise harvesting. However, existing banana detection and positioning methods have two main limitations in practical applications: large number of model parameters, difficult to deploy, and performance needs to be improved. To tackle the above issues, a high-precision and lightweight banana bunch recognition and positioning method was proposed and deployed on edge devices for application. First, a Slim-Banana model was proposed based on the improvement of YOLOv8l. In order to reduce the model calculation amount and maintain high performance, GSConv was introduced in the Slim-Banana model to replace the standard convolution, and combined with grouped convolution and spatial convolution. At the same time, the cross-stage local network (GSCSP) module was designed to reduce the computational complexity and the complexity of the network structure through a single-stage aggregation method. Then, the RealSense depth sensor is combined to perform image registration and 3D positioning of the banana using TOF technology. Finally, the pipeline is deployed on the Nvidia Orin NX edge device and its performance and resource consumption in actual work are deeply analyzed. Experimental results show that the detection precision, recall, mAP and inference time of our method are 0.947, 0.948, 0.98 and 113.6ms respectively, the network memory size required is 4449MiB, and the average positioning errors in the X-axis, Y-axis and Z-axis directions are 13.47mm, 12.87mm and 13.87mm respectively. To our knowledge, this is the first work that implements banana detection and localization on edge devices. Experimental results show that compared with existing methods, our method achieves better performance in complex orchard environments, achieving efficient and lightweight banana recognition and positioning.

Zitieren

APA

Zheng, Z., Chen, L., Wei, L., Huang, W., Du, D., Qin, G., Yang, Z., & Wang, S. (2024). An efficient and lightweight banana detection and localization system based on deep CNNs for agricultural robots. Smart Agricultural Technology, 9, 100550. https://doi.org/10.1016/j.atech.2024.100550

Vancouver

Zheng Z, Chen L, Wei L, Huang W, Du D, Qin G, et al. An efficient and lightweight banana detection and localization system based on deep CNNs for agricultural robots. Smart Agricultural Technology. 2024;9:100550. doi:10.1016/j.atech.2024.100550

BibTeX
@article{Zheng2024,
  author = {Zheng, Zhenhui and Chen, Ling and Wei, Lijiao and Huang, Weihua and Du, Dongjie and Qin, Guoquan and Yang, Zhou and Wang, Shuo},
  title = {An efficient and lightweight banana detection and localization system based on deep CNNs for agricultural robots},
  journal = {Smart Agricultural Technology},
  year = {2024},
  volume = {9},
  pages = {100550},
  doi = {10.1016/j.atech.2024.100550}
}
Kostenlos starten →