2007 Fiscal Year Final Research Report Summary
Studies of Cherry Harvesting by a Robot
| Project/Area Number |
17380154
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (B)
|
| Allocation Type | Single-year Grants |
|---|
| Section | 一般 |
|---|
| Research Field |
Agricultural information engineering
|
|---|
| Research Institution | Yamagata University |
|---|
Principal Investigator |
AKASE Akika Yamagata University, Faculty of Agriculture, Professor (30007075)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
FUJIURA Tateshi Osaka Prefecture University, Graduate School of Life and Environmental Sciences, Professor (00026585)
IMAGAWA Junichi Nara Prefectural Agricultural Experiment Station, Nara Fruit Tree Research Center, Chief Researcher (70393402)
|
|---|
| Project Period (FY) |
2005 – 2007
|
| Keywords | trellis cultivation / 3-D vision sensor / rectangular robot / harvesting hand / percentage of fruit harvestin / single trunk training / onenable fingers / low trunk |
|---|
| Research Abstract |
We manufactured a rectangular robot with a 3-D vision sensor for harvesting cherries grown by trellis cultivation. We made an opening-and-shutting type, a parallel link type, a ring type and a cut type as harvesting hands. The most suitable distance between a fruit and the robot was 30 cm. In the field test the hand could not always reach position of a peduncle. The gap was about 1 cm in maximum. Probability of suitable positioning of the hand was 57.2%. Percentage of fruit harvesting was over 80% in the opening-and-shutting type hand and the parallel link type hand at suitable positioning of the hand. In the ring type hand that was 0%, for the diameter of the ring was too small.
Another harvesting robot was manufactured for application in an orchard where cherry trees are cultivated using single trunk training in rows and are trained vertically with their branches pruned. The main parts of the robot are a manipulator with 4 degrees of freedom (DOF), a 3-D vision sensor, a hand, a compu
… More
ter and a traveling device. By processing the images from the 3-D vision sensor, the locations of the fruits and obstacles were recognized, and the trajectory of the hand was determined. Fruits were picked by the hand, while avoiding collisions with obstacles. An hand that sucks the fruit and grasps the peduncle to pick up was tested. The result showed that another fruit close to the target fruit is easy to be damaged by the grasping motion of the fingers. To prevent the damage, the distance between fingers was narrowed down. After decreasing the distance, the fruits were not damaged.
In the low single trunk training, it was able to get a yield of 500kg per 10 a in the fifth year after planting and 800kg per 10 a in the seventh year. In this training, the time of harvest was two-thirds compared with the vase form training, the time of pruning was one-fourth, and the time of pinching was three times. For harvesting in the middle of May, it was effective to set a minimum temperature at the start as 5℃ just after dormancy awakening in forcing-cultured cherry. Less
|
-
[Journal Article] Cherry-harvesting robot2008
Author(s)
Kanae, Tanigaki, Tateshi, Fujiura, Akira, Akase, Junichi, Imagawa
-
Journal Title
Computers and Electronics in Agriculture 63
Pages: 65-72
Description
「研究成果報告書概要(和文)」より
Peer Reviewed
-
-
-
-
-
-
-
-
-
