仿形(计算机编程)
环境科学
计算机科学
工程类
操作系统
作者
Antonio Ruiz-Gonzalez,Harriet Kempson,Jim Haseloff
出处
期刊:Micromachines
[MDPI AG]
日期:2024-10-24
卷期号:15 (11): 1293-1293
被引量:1
摘要
The development of low-cost tools for rapid soil assessment has become a crucial field due to the increasing demands in food production and carbon storage. However, current methods for soil evaluation are costly and cannot provide enough information about the quality of samples. This work reports for the first time a low-cost 3D printed device that can be used for soil classification as well as the study of biological activity. The system incorporated multiple physical and gas sensors for the characterisation of sample types and profiling of soil volatilome. Sensing data were obtained from 31 variables, including 18 individual light wavelengths that could be used to determine seed germination rates of tomato plants. A machine learning algorithm was trained using the data obtained by characterising 75 different soil samples. The algorithm could predict seed germination rates with high accuracy (RSMLE = 0.01, and R
科研通智能强力驱动
Strongly Powered by AbleSci AI