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

Fundamental research far an intelligent plant factory based on the computerized network.

基于计算机网络的智能植物工厂的基础研究。

基本信息

批准号：
11460121
负责人：
HASHIMOTO Yasushi
金额：
$ 9.66万
依托单位：
Ehime University
依托单位国家：
日本
项目类别：
Grant-in-Aid for Scientific Research (B)
财政年份：
1999
资助国家：
日本
起止时间：
1999 至 2001
项目状态：
已结题

项目摘要

In this study, an effective data communication technique for a decentralized plant production system consisting of many plant factories, which works on the network, was explored.1. Image data are characterized by large size. An effective image compression technique for plant was developed to deal well with such large sized data on the network. It was applied for diagnosing the growth conditions of plants cultivated in different plant factories by transmitting their plant images. This technique was very useful for data communication among several plant factory systems. (Paper 1)2. A stereo sensing technique was developed to construct 3-dimensional plant shape in real-time processing. A fruit image was first obtained by composing two images took from the right and left directions using two CCD cameras and then extracted its outline by displaying with linear lines. From these procedures, we obtained a 3-dimensional plant model. This technique significantly shortened the processing time fo … More r 3-D image construction. We could obtain a more effective 3-dimentional model, which looks like an image taken by a real human vision, by using the concept of a tele-existence method. (Papers 2 and 3)3. An image recognition technique for a visual sensor of a harvesting robot was developed and applied it to the evaluation of the quality of seedlings which are moving on the belt conveyor. This technique provided a high-speed processing (5 images/s) for evaluation. This technique was useful to achieve automation of a plant evaluation system. (Paper 4)4. Shape evaluation of fruit is quite empirical and uncertain. A new technique was developed to evaluate the fruit shape quantitatively using attractor, fractal dimension and neural networks. There was high correlations among identification errors, the shapes of attractors and fractal dimensions for evaluating fruit shape. So, these approaches allow the quantitative evaluation of the complexity of the fruit shape and are useful for the data communication among plant factories. (Paper 5)5. Finally, total mechanization and automation for decentralized plant factories were investigated using newly developed image processing techniques mentioned above. (Paper 6) Less

本研究针对多个植物工厂组成的分散式植物生产系统，探讨一种有效的数据通讯技术.图像数据的特点是数据量大。为了处理网络上大量的植物图像数据，提出了一种有效的植物图像压缩技术。该系统通过传输植物图像来诊断不同植物工厂栽培的植物的生长状况。该技术对于多个工厂系统之间的数据通信非常有用。（论文1）2.提出了一种立体感测技术，用于实时处理植物的三维形状。首先利用两个CCD摄像机从左右两个方向拍摄的图像合成得到水果图像，然后用直线显示提取水果的轮廓。通过这些步骤，我们得到了一个三维植物模型。这项技术大大缩短了处理时间， ...更多信息 r三维图像构建。通过使用远程存在方法的概念，我们可以获得一个更有效的三维模型，它看起来像一个真实的人类视觉所拍摄的图像。（第2、3页）提出了一种用于采集机器人视觉传感器的图像识别技术，并将其应用于带式输送机上秧苗质量的评价。该技术提供了用于评价的高速处理（5个图像/s）。该技术有助于实现植物评价系统的自动化。（论文4）4.果实形状的评价具有很强的经验性和不确定性。提出了一种利用吸引子、分形维数和神经网络对果实形状进行定量评价的新方法。果实形状的识别误差、吸引子形状和分形维数三者之间存在高度相关性。因此，这些方法允许的水果形状的复杂性的定量评估，并为植物工厂之间的数据通信是有用的。（论文5）5.最后，使用上述新开发的图像处理技术，对分散式工厂的完全机械化和自动化进行了研究。（纸6）减