权益分类	功能权益	普通用户	{{item.name}}会员
{{category.name}}	{{benefitItem.name}}

Reduction of Agrcultural Chemicals by Automatic Pest Control

通过自动害虫防治减少农业化学品

基本信息

批准号：
06454106
负责人：
OKAMOTO Tsuguo
金额：
$ 4.54万
依托单位：
The University of Tokyo
依托单位国家：
日本
项目类别：
Grant-in-Aid for Scientific Research (B)
财政年份：
1994
资助国家：
日本
起止时间：
1994 至 1996
项目状态：
已结题

项目摘要

[1]To develop an auto-mcchanized future system of weeding, we investigated a new method to reduee the quantity of herbicide application. In the scrial experiment on a ficld practieed from April to December in 1995, we carried out the factorial expcriment composed of kind of weed (four levels) * kind of herbicide (three level) * application time (three levels) * application methods (three levels) * concentration of chemical (three levels) * repctition (ten plants). Although the weeding effect by conventional spraying method (control) was gencrally higher than that by other methods (painting and injeeting), the high effects to reduce the quantity of application were shown in the trial methods. If we select the way to give an only small damage to weed in place of killing weed in perfect, our proposed methods are thought to be useful in future with further investigation.[2]This study describes the image gridance for autonomous agricultural vchicle for mechanical weeding. Image analysis was used for an autonomous agricultural vchiclc, and the TV camcra direction to the furrow and crop row was identified from the ficld images. Principlc component analysis and HSI transformation were applied to the image binarization. The boundary of furrow and crop row was approximated by two lines using Hough transform and the camcra position was idcntified from the parameter of these two lines. The results showed HSI transform to be exccllent, and the estimated position to be in good agreement with the actual position.

[1]To通过研制一种自动化除草机，探索了一种减少除草剂用量的新方法。在1995年4 ~ 12月的田间试验中，采用杂草种类（4个水平）×除草剂种类（3个水平）×施药时间（3个水平）×施药方法（3个水平）×药剂浓度（3个水平）×处理（10株）的析因试验。常规喷雾法（对照）的除草效果一般高于其它方法（涂抹和注射），但试验方法也显示出减少施药量的效果。如果我们选择的方式，给一个只有很小的伤害杂草，而不是杀死杂草在完美的，我们提出的方法被认为是有用的，在未来的进一步调查。[2]本研究描述了用于机械除草的自主农业车辆的图像网格。本文对一台自主式农用拖拉机进行了图像分析，从田间图像中识别出电视摄像机对犁沟和作物行的方向。利用主成分分析和HSI变换对图像进行二值化。利用Hough变换将垄行边界近似为两条直线，并根据这两条直线的参数识别垄位。结果表明，HSI变换效果良好，估计位置与实际位置吻合较好。