基本信息:
- 专利标题: 土の粒度分布の推定方法
- 专利标题(英):METHOD FOR ESTIMATING PARTICLE SIZE DISTRIBUTION OF SOIL
- 申请号:JP2020009579 申请日:2020-01-24
- 公开(公告)号:JP2021117625A 公开(公告)日:2021-08-10
- 发明人: 岩下 将也 , 大塚 義一
- 申请人: 株式会社奥村組
- 申请人地址: 大阪府大阪市阿倍野区松崎町2丁目2番2号
- 专利权人: 株式会社奥村組
- 当前专利权人: 株式会社奥村組
- 当前专利权人地址: 大阪府大阪市阿倍野区松崎町2丁目2番2号
- 代理人: 特許業務法人翔和国際特許事務所
- 主分类号: G06N20/00
- IPC分类号: G06N20/00 ; G06T7/00 ; G06T7/62
To provide a method for estimating particle size distribution of soil capable of accurately estimating the particle size distribution of soil containing small particles by machine learning with artificial intelligence.SOLUTION: Particle size distribution of collected soil is estimated with a trained model 20 (see Fig. 3) that is obtained by machine learning with artificial intelligence built into a server 11 using, as training data, a large number of images of soil in each single particle size that was screened from raw material soil and classified into a plurality of particle sizes. The images of soil in each single particle size used as the training data is an image of the surface spread with a predetermined thickness T that is captured by a photographing device 12 and linked with each classified particle size. The image, which is obtained by capturing the image of the surface of the collected soil spread with the predetermined thickness T by the predetermined photographing device 12, is input to the trained model 20 to estimate the particle size distribution of the collected soil by outputting a ratio of each classified particle size of the collected soil.SELECTED DRAWING: Figure 1