2001 Fiscal Year Final Research Report Summary
Automated mechanical weed control aiming sustainable agriculture
| Project/Area Number |
11660258
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Single-year Grants |
|---|
| Section | 一般 |
|---|
| Research Field |
農業機械学
|
|---|
| Research Institution | Kagoshima University |
|---|
Principal Investigator |
IWASAKI Koichi Kagoshima University, Faculty of Agriculture, Associate Professor, 農学部, 助教授 (90232652)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
SUEYOSHI Takeshi Kagoshima University, Faculty of Agriculture, Research Associate, 農学部, 助手 (40305165)
MIYABE Yoshiteru Kagoshima University, Faculty of Agriculture, Professor, 農学部, 教授 (30041643)
|
|---|
| Project Period (FY) |
1999 – 2001
|
| Keywords | Sustainable Agriculture / Weed control / Recycled paper / Used paper / Machine Vision |
|---|
| Research Abstract |
Development of weed control system without agricultural chemical is expected for sustainable agriculture. To establish the weed suppression system using paper slurry, influences of the level of pulverizing and the quantity of paper slurry to be applied were examined.
(1) Radish seeds were used and following results were obtained. 1) The paper slurry with medium pulverizing showed good effect of suppressing the emergence when more than 250g/m^2 was applied. 2) The paper slurry with fine, medium and rough pulverizing showed effect on suppression the emergence when more than 250g/m^2 were applied.
(2) With these results, weeds suppression test in the field was carried out using medium pulverized paper slurry and the results are as follows. 1) Paper slurry is effective for preventing increase of soil temperature. 2) The number of weed emergence was less than 1/4 of those in area with no paper slurry applied. And the more the paper slurry applied, the more effective on weed suppression. 3) The dried weight of weeds at the end of field test showed similar results to the emerged number of weeds. 4) To determine the quantity of paper slurry to be applied, desired period of weed suppression and allowable weeds should be considered.
(3) To develop a judgment system of weeds and plants using a machine vision and neural networks algorithm, shape judgment software was composed by video capture control, image processing, feature extraction and neural networks judgment routines. It was developed using Microsoft Visual Basic language. From the result of judgment using this system, the accuracy of recognition of plants , was significantly good.
|
