All publications from your selection

Download RAW file:
DOI:
Description: Soil particle size distribution (soil texture) is an important determinant of agronomic practices, such as field management, crop selection, nutrient management, and irrigation scheduling. However, the current methods for measuring soil texture are time-consuming and/or expensive. The prevailing hydrometer method relies on meticulous wet chemistry carried out at specialized laboratories on soil samples laboriously collected, labeled, and shipped from the field. Furthermore, alternative approaches based on infrared spectroscopy rely on expensive equipment. This paper presents a new, proximal sensing method for soil texture analysis which combines image acquisition via a low-cost digital microscope with computer vision algorithms to enable inexpensive and rapid determination of particle size distribution. Images may be collected in or near the field and uploaded for processing in the cloud. Unlike earlier digital microscopy methods, the new deep-learning algorithm employs a convolutional neural network (CNN) model trained to detect features of soil images which correlate with soil texture. Training is performed via supervised machine learning guided by hundreds of sample images labeled with their corresponding hydrometer-based soil texture measurements.

Keywords:
Résumé:
Mots-clés:
Citation:
Volume:
Issue:
Pages -
Contributor:
Date: 2021-06-11
Technical field:
Conference name: 5th CIGR International Conference and CSBE-SCGAB AGM 2021, Quebec City,QC, 11-14 May 2021.
Session name: World Congress on Computers in Agriculture and Natural Resources (WCCA 2)

Other information:
Type: Text.Abstract
Format:
Publication type: Presentation
Source:
Relation:
Coverage: Canada
Language 1: en
Language 2:
Rights: Canadian Society for Bioengineering
Notes:
Other files: