| Online detection of tomatoes for robotic harvesting |
| Paper ID : 1405-NICAME1402 |
| Authors: |
|
Hossein Behfar *1, Amin Helmi2, Delara Ghanini3 1Department of Biosystems Engineering, University of Tabriz, Tabriz, Iran 2گروه مهندسی بیوسیستم، دانشکده کشاورزی، دانشگاه تبریز، تبریز، آذربایجان شرقی، ایرا ن 3Department of Biosystem Engineering, Faculty of Agriculture, Tabriz University, Tabriz, East Azarbaijan, Iran |
| Abstract: |
| Harvesting of greenhouse crops such as tomatoes is considered one of the most important operations of agricultural systems, because a high-quality product is harvested at the optimal time, with minimal damage. In order to improve and optimize this process, robotic harvesting methods have been proposed as a new solution. Since in these methods, online detection of tomatoes is of particular importance, robots can automatically identify and harvest tomatoes. Agricultural robots are sensitive and intelligent machines that are programmed to perform various agricultural tasks such as planting, transplanting, spraying, pruning and harvesting. One of the main reasons for agricultural automation is the reduction of manpower, and for this reason, automatic and robotic harvesting are particularly important in agricultural automation. This article deals with the design, simulation and control of six degrees of freedom robotic arm. In this research, the three-axis platform is designed and built so that it can be controlled using the image processing control system and perform various operations such as harvesting on the product with appropriate operators. Hopefully, after detecting the position of the target (which in this paper is a tomato), the robot's trailing arm will be directed towards it. The accuracy of this system will be evaluated in approaching the product. This article examines different methods of online detection of tomatoes for robotic harvesting.Methods based on shape features, color features and deep learning are among the approaches that are investigated in this paper. |
| Keywords: |
| Robot, on-line detection, automation, image processing, control |
| Status : Paper Accepted (Poster Presentation) |
