Computer Vision-Assisted Object Detection and Handling Framework for Robotic Arm Design Using YOLOV5
ADCAIJ: Advances in Distributed Computing and Artificial Intelligence Journal · 29.12.2023 · 12 · e31586
DOI: 10.14201/adcaij.31586 ↗Kurz gefasst
Fachartikel in der Zeitschrift ADCAIJ: Advances in Distributed Computing and Artificial Intelligence Journal, online erschienen am 29.12.2023. Verfasst von Ajmisha Maideen und 1 weiteren Autor:innen. Die Arbeit wurde bisher 5-mal zitiert (Crossref) und nennt 21 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC-ND 4.0).
Abstract (Originalsprache)
In recent years, there has been a surge in scientific research using computer vision and robots for precision agriculture. Productivity has increased significantly, and the need for human labor in agriculture has been dramatically reduced owing to technological and mechanical advancements. However, most current apple identification algorithms cannot distinguish between green and red apples on a diverse agricultural field, obscured by tree branches and other apples. A novel and practical target detection approach for robots, using the YOLOV5 framework is presented, in line with the need to recognize apples automatically. Robotic end effectors have been integrated into a Raspberry Pi 4B computer, where the YOLOV5 model has been trained, tested, and deployed. The image was taken with an 8-megapixel camera that uses the camera serial interface (CSI) protocol. To speed up the model creation process, researchers use a graphical processing computer to label and preprocess test images before utilizing them. Using YOLOV5, a computer vision system-assisted framework aids in the design of robotic arms capable of detecting and manipulating objects. The deployed model has performed very well on both red and green apples, with ROC values of 0.98 and 0.9488, respectively. The developed model has achieved a high F1 score with 91.43 for green apples and 89.95 for red apples. The experimental findings showed that robotics are at the forefront of technological advancement because of the rising need for productivity, eliminating monotonous work, and protecting the operator and the environment. The same discerning can be applied to agricultural robots, which have the potential to improve productivity, safety, and profit margins for farmers while reducing their impact on the environment. The system’s potential could be seen in an assortment of fields, including sophisticated object detection, nuanced manipulation, multi-robot collaboration, and field deployment.
Zitieren
APA
Maideen, A., & Mohanarathinam, A. (2023). Computer Vision-Assisted Object Detection and Handling Framework for Robotic Arm Design Using YOLOV5. ADCAIJ: Advances in Distributed Computing and Artificial Intelligence Journal, 12, e31586. https://doi.org/10.14201/adcaij.31586
Vancouver
Maideen A, Mohanarathinam A. Computer Vision-Assisted Object Detection and Handling Framework for Robotic Arm Design Using YOLOV5. ADCAIJ: Advances in Distributed Computing and Artificial Intelligence Journal. 2023;12:e31586. doi:10.14201/adcaij.31586
BibTeX
@article{Maideen2023,
author = {Maideen, Ajmisha and Mohanarathinam, A},
title = {Computer Vision-Assisted Object Detection and Handling Framework for Robotic Arm Design Using YOLOV5},
journal = {ADCAIJ: Advances in Distributed Computing and Artificial Intelligence Journal},
year = {2023},
volume = {12},
pages = {e31586},
doi = {10.14201/adcaij.31586}
}
EUVIMED