Application of Image Recognition Technology in Target Recognition of Automated Apple Picking Robot

  • Ya Wang

Abstract

Scientific and reasonable flower and fruit management of apple trees is conducive to
ensuring the health of fruit trees, so as to achieve high quality and high yield of apples.
Different varieties of apples have different fruit types and different fruit setting rates.
Therefore, the picking robot needs to install a flexible end effector to ensure the integrity
of the picking fruit. Different fruits will grow in different locations. For example, some
fruits grow on trees and some fruits grow on vines. In addition, there are differences in
the size and maturity of the fruit, so the robot's visual positioning is more difficult. When
designing a picking manipulator, the cultivation methods of fruits and vegetables should
be considered to ensure that the manipulator can meet the picking requirements, and will
not damage the stems, leaves and other parts of the plant, and can accurately grasp the
fruits according to the set route. Therefore, manipulators need to have the ability to avoid
obstacles, and redundant manipulators can be used. This kind of mechanical means has
obvious advantages, but because of the higher degree of freedom, the control is more
difficult. Picking robots are mainly operated by farmers, so the operation of the robots
should not be too difficult to ensure that farmers can quickly learn and master. In
addition, the production cost of robots must be controlled to ensure that farmers can
afford the cost of purchasing robots.

Published
2020-11-01
Section
Articles