Machine learning and computer vision technology to analyze and discriminate soil samples


Kaplan S., Ropelewska E., Günaydın S., Sabancı K., Çetin N.

SCIENTIFIC REPORTS, cilt.14, sa.19945, ss.1-11, 2024 (SCI-Expanded) identifier

  • Yayın Türü: Makale / Tam Makale
  • Cilt numarası: 14 Sayı: 19945
  • Basım Tarihi: 2024
  • Doi Numarası: 10.1038/s41598-024-69464-7
  • Dergi Adı: SCIENTIFIC REPORTS
  • Derginin Tarandığı İndeksler: Science Citation Index Expanded (SCI-EXPANDED), Scopus, Academic Search Premier, BIOSIS, Chemical Abstracts Core, MEDLINE, Veterinary Science Database, Directory of Open Access Journals
  • Sayfa Sayıları: ss.1-11
  • Ankara Üniversitesi Adresli: Evet

Özet

Soil texture is one of the most important elements to consider before planting and tillage. Thesefeatures affect the product selection and regulate its water permeability. Discrimination of soils bydetermining soil texture features requires an intense workload and is time-consuming. Therefore,having a powerful tool and knowledge for texture-based soil discrimination could enable rapid andaccurate discrimination of soils. This study focuses on presenting new models for 6 different soilsample groups (Soil_1 to Soil_6) based on 12 different machine learning algorithms that can beutilized for various problems. As a result, overall accuracy values were determined as greater than99.2% (Trilayered Neural Network). The greatest accuracy value was found in Bayes Net (99.83%)and followed by Subspace Discriminant (99.80%). In the Bayes Net algorithm, MCC (MatthewsCorrelation Coefficient) and F-measure values were obtained as 0.994 and 0.995 for Soil_4 and Soil_6sample groups while these values were 1.000 for other soil groups. Soil types can visually vary basedon their texture, mineral composition, and moisture levels. The variability of this can be influencedby fertilization, precipitation levels, and soil cultivation. It is important to capture the images in soilconditions that are more stable. In conclusion, the present study has proven the feasibility of rapid,non-destructive, and accurate discrimination of soils by image processing-based machine learning.