O. I. Mikhailovskaya

O. I. Mikhailovskaya

Auto-generated profile, based on publicly available information

O. I. Mikhailovskaya

Anapa Zonal Experimental Station of Viticulture and Winemaking is a branch of the Federal State Budgetary Scientific Institution “North Caucasian Federal Scientific Center of Horticulture, Viticulture and Winemaking

Research Focus & Metrics

Academic Career Length

3Years (2024-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
24 October 2024 in Agrarian science
Verified
Journal Article
Open access
PDF file
VOV. A. Orlov
V. A. Orlov
LALuk'ianov Aleksei
Luk'ianov Aleksei
OMO. I. Mikhailovskaya
O. I. Mikhailovskaya
Introduction. Soils play an important role in the approximately 30-year period of operation of a grape planting, influencing plant growth, their yield and the quality of the grapes. In this study, the morphometric parameters of the surface soil layer of a grape plantation were studied using spectral channels of satellite images.Methodology. The methodology included the use of a “random forest” algorithm to classify soil cover using spectral channels and normalized satellite image indices and analyze the main physicochemical properties of soils. Accuracy was assessed using RMSD and confidence intervals calculated via bootstrapping.Results. The study revealed significant differences in the spectral reflectivity of different site options, which was due to carbonate content, humidity levels and the amount of humus. Areas with high carbonate and moisture content showed higher standard deviation values in the spectral channels. Studying the spectral characteristics of the soil surface makes it possible to effectively classify different areas based on remote sensing data. Analysis of combinations of spectral channels revealed an optimal set of three channels (B12, B11, B8A) with a minimum standard deviation when classifying an image into six soil variants of areas. For classification, a composition of five normalized indices can also be used, but in this case the calculation time increases significantly with a larger standard deviation and a larger confidence interval range. Using machine learning, six distinct soil surface types were segmented, demonstrating the complexity of the field›s soil mosaic. These results are critical for improving vineyard management and productivity