Picking is one of the most important aspects in the production of fruits and vegetables and directly affects the market value of fruits and vegetables. Since the first tomato-picking robot was born in the United States in 1983, the research and development of picking robots have been going on for more than 20 years. Japan, Europe and the United States and other countries have successively researched and picked intelligent robots such as apples, oranges, tomatoes, watermelons and grapes.
At present, Japan has done a lot of research in the field of fruit picking robots. The picking robots it researches mainly include the following types.
1. Tomato Picking Robot Japan's Kondo-N et al. developed a tomato harvesting robot consisting of a manipulator, an end effector, a visual sensor, and a moving mechanism. A cluster of tomatoes can grow 4-6 fruits. Each fruit does not necessarily mature at the same time, and the fruit is sometimes blocked by leaf stems. During the harvest, it requires a large range of robot movements to avoid obstacles, so the robot picking robot is designed to have 7 Degrees of freedom, able to pick a designated picking posture for picking. The end effector is composed of two mechanical fingers and a sucker; the visual sensor mainly uses a color camera to find and identify mature fruits, and uses the binocular vision method to locate the target; the moving mechanism adopts a four-wheel structure and can automatically walk between ridges. During picking, the moving mechanism walks a certain distance and then performs image acquisition. The visual system is used to detect the position information of the fruit relative to the robot coordinate system, to judge whether or not the tomatoes are within the harvesting range. If it is possible to harvest, then the control robot is approached and picked. Take the fruit, the suction cup sucks the fruit, the mechanical finger catches the fruit, and then unscrew the fruit through the wrist of the robot.
2. Strawberry picking robot Kondo-N et al. also developed corresponding picking robots for different cultivation modes of strawberry (overhead cultivation mode and traditional mode). Overhead cultivation patterns are increasingly used due to robotic operations. The robot uses a 5-DOF picking robot. The vision system is similar to the tomato picking robot. The end effector uses a vacuum system and a screw-accelerated cutter. At the time of harvest, the visual system calculates the spatial position of the picking target, then the picking robot moves to a predetermined position, the end effector moves downwards until the strawberry is sucked in; the position of the strawberries is detected by 3 pairs of photoelectric switches, when the strawberries are in the right position, When the wrist moves, the fruit stem enters the designated position, and the spiral acceleration drives the cutter to cut off the fruit stem to complete picking.
3. Cucumber Picking Robot The cucumber picking robot uses a 6-DOF robot and can work under a tilted shelf. This type of bracket cultivation is specifically designed for mechanized picking. The cucumber fruit is on the underside of the sloped shed, which facilitates the separation of cucumbers from stems and leaves, making detection and picking easier. A filter was added in front of the camera to identify the cucumber based on the spectral reflection characteristics of the cucumber. The end effector is equipped with a stem detector, a cutter and a mechanical finger. When the cucumber is picked by a mechanical finger during picking, the stem detector finds the stem, and then the cutter cuts the stem.
4. Multi-function grape picking robot The grape picking robot uses a 5-DOF polar robot and the end arm can move at a uniform speed under the trellis. The visual sensor generally adopts the color video camera, uses the PSD three-dimensional vision sensor to be better, may examine the three-dimensional information of the ripe fruit and the distance information. In the open planting mode, because the picking season is too short and the single picking function makes the use efficiency of the robot too low, a variety of end effectors have been developed, such as end effectors for picking and bagging, respectively. The nozzle at the end of the robot, etc. The end effector for grape harvesting has mechanical fingers and scissors. When picking, the fruit house is grasped with a mechanical finger and the handle is cut with scissors. In addition to the several types of picking robots described above, Japan has also developed robots for citrus picking, mushroom and watermelon harvesting, and the like. At present, the intelligence level of picking robots for fruits and vegetables is still limited, and there is still a certain distance from practicality and commercialization. The main problems are: first, the fruit recognition rate and picking rate are not high, and the damage rate is high; second, the average picking cycle of the fruit is longer; third, the picking robot has higher manufacturing cost; along with sensors and computer vision technologies, etc. To develop, the research on picking robots for fruits and vegetables needs to be carried out in the following aspects: First, it is necessary to find a reliable and high-precision vision system technology that can detect all mature fruits and precisely locate them; second, improve the robots. And the end-effector design flexibility and dexterity, successful obstacle avoidance, improve the success rate of picking, reduce the damage rate of the fruit; Third, to improve the versatility of picking robots, improve the utilization of the robot.
ZHEJIANG EVERNEW SEAFOOD CO.,LTD , https://www.evernewseafood.com