thumbnail

Topic

Technologies and technical equipment for agriculture and food industry

Volume

Volume 74 / No. 3 / 2024

Pages : 837-845

Metrics

Volume viewed 0 times

Volume downloaded 0 times

DEEP LEARNING PREDICTIVE MODEL FOR SOIL TEXTURAL ASSESSMENT

PAG TUKOY SA URI NG LUPA GAMIT ANG MODELO NG DEEP LEARNING

DOI : https://doi.org/10.35633/inmateh-74-74

Authors

KARLA JANE QUINOL

Central Luzon State University / Philippines

(*) CAROLYN GRACE SOMERA

Central Luzon State University / Philippines

MARVIN CINENSE

Central Luzon State University / Philippines

NEMESIO JR. MACABALE

Central Luzon State University / Philippines

(*) Corresponding authors:

[email protected] |

CAROLYN GRACE SOMERA

Abstract

The distribution of grain sizes in different soil samples is essential for agriculture and geotechnics, providing high-resolution soil maps crucial for land use planning. Traditional methods for soil texture analysis are reliable but often time-consuming and inconsistent. With that, this study aims to create an efficient predictive model for soil texture classification using deep learning techniques. A dataset of 4,556 images was extensively pre-processed and trained, with a model chosen for validation due to its low MSE value of 1.18. The model's performance, evaluated through Precision, Recall, and F1 Score, showed weighted averages of 88%, 78%, and 74%, respectively, and an overall accuracy of 94.56%. Validation using 456 images revealed high accuracy for Sandy and Clayey Soils but varying results for Loamy and Silty Soils. In Trial 1, the model achieved over 91% accuracy for all soil textures, with 100% accuracy for Sandy Soil. However, Trials 2 and 3 exhibited decreased accuracy for Loamy and Silty Soils, with the lowest accuracies at 61.40% and 65.78%, respectively. These results suggest that while the model is effective for certain soil textures, it requires further refinement and additional diverse training data to consistently match the reliability of traditional methods.

Abstract in Tagalog

Ang pagtukoy sa uri ng lupa ay mahalaga sa larangan ng agrikultura at geotechnics. Ito ang nagbibigay ng maayos na mapa na siyang kritikal sa pagpaplano ng paggamit ng lupa. Ang mga tradisyunal na pamamaraan sa pagtukoy nito ay maaasahan, ngunit kadalasang matagal ang proseso at hindi pare-pareho. Dahil dito, ang pagsusuring ito ay naglalayong lumikha ng mabisang modelo para sa klasipikasyon ng uri ng lupa gamit ang makabagong teknolohiya na deep learning. Ang dataset na may 4,556 imahe ay sumailalim sa pag-proproseso, bago ginamit sa paghasa ng iba’t ibang modelo, kung saan ang napili para sa balidasyon ay may mababang MSE value na 1.18. Ang bisa ng modelo na sinukat sa pamamagitan ng Precision, Recall, at F1 Score, ay nagpakita ng mga weighted average na 88%, 78%, at 74%, at may kabuuang accuracy naman na 94.56%. Sa balidasyon gamit ang 456 imahe, ipinakita ang mataas na accuracy para sa Sandy (Mabuhangin) at Clayey (Luwad) na lupa ngunit may iba't ibang resulta para sa Silty (Maalikabok) at Loamy (kumbinasyon ng tatlo) na lupa. Sa unang eksperimento, nakamit ng modelo ang 91% accuracy para sa lahat ng uri ng lupa, na may 100% accuracy para sa Sandy soil. Gayunpaman, ang ikalawa at ikatlong eksperimento ay nagpakita ng pagbaba ng accuracy para sa Loamy (61.40%) at Silty (65.78%) Soils. Ipinahihiwatig nito na habang ang modelo ay epektibo sa ilang uri ng lupa, kailangan pa itong mapabuti at dagdagan ng mas magkakaibang datos sa pag-hasa upang ganap na maitatag ang pagiging maaasahan nito.

Indexed in

Clarivate Analytics.
 Emerging Sources Citation Index
Scopus/Elsevier
Google Scholar
Crossref
Road