ارزیابی انحراف ربات سمپاش خودکار خورشیدی از مسیر آموزش داده شده در سرعت‌ها و سطوح متفاوت

نوع مقاله : مقاله پژوهشی

نویسندگان

• علیرضا احمدلو ¹
• محمد عسکری ²
• داود کلانتری ³

¹ دانشجوی کارشناسی ارشد، گروه مهندسی مکانیک بیوسیستم، دانشکده مهندسی زراعی، دانشگاه علوم کشاورزی و منابع طبیعی ساری، ساری، ایران

² استادیار گروه مهندسی مکانیک بیوسیستم، دانشکده مهندسی زراعی، دانشگاه علوم کشاورزی و منابع طبیعی ساری

³ دانشیار گروه مهندسی مکانیک بیوسیستم، دانشکده مهندسی زراعی، دانشگاه علوم کشاورزی و منابع طبیعی ساری، ساری، ایران

چکیده

کلیدواژه‌ها

• سمپاش باغی
• تصحیح مسیر هوشمند
• رباتیک
• مسیریابی
• YOLO v8

موضوعات

• رباتیک و اتوماسیون

عنوان مقاله [English]

Evaluation of deviation of solar-powered autonomous sprayer robot from the trained path at different speeds and surfaces

نویسندگان [English]

• Alireza Ahmadlou ¹
• mohammad askari ²
• Davood kalantari ³

کلیدواژه‌ها [English]

• Garden sprayer
• smart route detection
• robotic
• routing
• YOLO v8

EXTENDED ABSTRACT

Introduction

The growing global demand for food, coupled with the need to support the emergence of technologies that can meet environmental standards as well as promote efficiency and healthy working environments, poses a major challenge for agriculture. Agricultural land is limited and can only be expanded slightly. Precision spraying represents an innovative approach that can help farmers spray only when needed, using the right amount of chemicals. Precision spraying can help meet regulatory requirements in terms of worker health, environmental sustainability, and food safety and quality. The best solution to this problem is to carry out pesticide spraying without human intervention and use a robotic system in agriculture. Agricultural robots are used to work in greenhouse and garden environments and play a very important role in agricultural activities. According to the survey, no research has been conducted so far on evaluating the deviation of an automated solar sprayer robot from the trained path at different levels and speeds.

Materials and Methods

The solar-powered automatic sprayer robot was equipped with accessories such as a chassis, wheels, electric motor, gearbox, solar panel, and power storage battery, etc. Field tests were conducted on two factors: surface type and forward speed. For this purpose, after training the route using the optimized YOLO v8 model for the solar-powered automatic sprayer robot, tests were conducted considering two input variables including three surface types (asphalt, concrete, and gravel soil) and three forward speeds (1, 2, and 3 km/h), and the amount of deviation of the robot from the trained route was measured as the output variable. The robot's deviation (in centimeters) from the original path was calculated with an assumed width of 50 centimeters.

Results and Discussion

Using morphological processing and thresholding methods, road lines were extracted and the upper and lower edge points of the road were determined using the contours of the largest white areas. Finally, red lines were drawn as the side lines of the road to identify the outer areas of the route. A yellow line was also drawn as the center line of the road as a navigation guide. The processed output images show that Gabor filters help identify and separate the road from the background by detecting edges and road structures. This filter also highlights the road lines and distinguishes them from other parts of the image, which is very important for robot navigation on uneven and varied paths. The YOLO model was used to identify road areas and create masks required for more accurate processing. The results showed that high speed and rough road surface are the two main factors in increasing the robot's deviation from the path. Analysis of the results of the YOLO v8 model in identifying garden paths indicates its high performance in accuracy, recall, and reduction of false positive and negative errors. The optimized YOLO v8 model, with its high accuracy and generalizability, showed very good performance in the navigation of the sprayer robot and is very suitable for practical applications such as automatic navigation in garden environments.

Conclusion

Image processing using Gabor filters and morphological processing methods helped improve the accuracy of the model and was able to distinguish the features of the paths well. Comparison of the results with previous research shows that the new model has significant improvements over other methods and is suitable for practical applications in the field of automatic navigation and garden path recognition. These findings can be useful for optimizing robot guidance algorithms in different environmental conditions and speeds.

Author Contributions

Conceptualization, A.A. and M.A.; methodology, A.A.; software, D.K.; validation, M.A., formal analysis, D.K.; investigation, M.A.; resources, A.A.; data curation, M.A.; writing—original draft preparation, M.A.; writing—review and editing, D.K.; visualization, A.A.; supervision, M.A.; project administration, M.A.; funding acquisition, M.A. All authors have read and agreed to the published version of the manuscript.

Data Availability Statement

Data available on request from the authors.

Acknowledgements

The authors would like to thank all participants of the present study.

Ethical considerations

The authors avoided data fabrication, falsification, plagiarism, and misconduct.

Conflict of interest

The author declares no conflict of interest.